added RAG, Multiuser, TG bot

.env.example

@@ -1,106 +1,121 @@
-# Server (internal bind inside containers)
-HOST=0.0.0.0
-BACKEND_INTERNAL_PORT=8080
-FRONTEND_INTERNAL_PORT=80
-
-# External ports on the host (docker compose publish)
-BACKEND_PORT=8080
-FRONTEND_PORT=3080
-VITE_DEV_PORT=5173
-
-# OpenRouter
-OPENROUTER_API_KEY=sk-or-v1-your-key-here
-OPENROUTER_MODEL=deepseek/deepseek-chat
-# deepseek/deepseek-v4-pro — сильная модель, tools поддерживаются:
-# OPENROUTER_MODEL=deepseek/deepseek-v4-pro
-OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
-OPENROUTER_TOOLS_ENABLED=true
-# none = без thinking (быстрее, стабильнее с tools). low|medium|high|xhigh — reasoning.
-OPENROUTER_REASONING_EFFORT=none
-# JSON-экстракция памяти отдельной моделью (если основная капризничает):
-# MEMORY_EXTRACT_MODEL=deepseek/deepseek-chat
-
-# App
-DATABASE_URL=sqlite:///./data/assistant.db
-CORS_ORIGINS=http://localhost:5173,http://localhost:8080,http://localhost:3080
-SYSTEM_PROMPT_PATH=./prompts/assistant.md
-MEMORY_AUTO_EXTRACT=true
-
-# Fitness (wger + Open Food Facts — public HTTPS, no proxy)
-WGER_BASE_URL=https://wger.de/api/v2
-OPENFOODFACTS_BASE_URL=https://world.openfoodfacts.org
-FITNESS_REMINDERS_ENABLED=true
-REMINDERS_ENABLED=true
-
-# Taiga (on host :9000, nginx → taiga.grigowashere.ru)
-TAIGA_BASE_URL=http://host.docker.internal:9000
-TAIGA_USERNAME=your_taiga_user
-TAIGA_PASSWORD=your_taiga_password
-TAIGA_PUBLIC_URL=https://taiga.grigowashere.ru
-
-# Gitea (on host :3000, nginx → git.grigowashere.ru)
-GITEA_BASE_URL=http://host.docker.internal:3000
-GITEA_TOKEN=your_gitea_api_token
-GITEA_PUBLIC_URL=https://git.grigowashere.ru
-GITEA_WEBHOOK_SECRET=generate_a_random_secret
-
-# Gitea webhook URL (repo Settings → Webhooks):
-# https://assistant.your-domain/api/v1/webhooks/gitea   ← nginx → 127.0.0.1:BACKEND_PORT
-# http://172.17.0.1:8202/api/v1/webhooks/gitea        ← если Gitea в Docker (не 127.0.0.1!)
-
-REPOS_DIR=/data/repos
-
-# Homelab — GPU PC 192.168.1.109
-OPENMETEO_BASE_URL=http://192.168.1.109:8085
-WEATHER_LAT=59.9343
-WEATHER_LON=30.3351
-WEATHER_LOCATION_NAME=Санкт-Петербург
-WEATHER_CACHE_SEC=300
-
-# News RSS (comma-separated)
-NEWS_RSS_URLS=https://habr.com/ru/rss/all/all/,https://www.reddit.com/r/programming/.rss
-NEWS_CACHE_SEC=1800
-NEWS_MAX_ITEMS=7
-
-# Morning digest (Europe/Moscow or user profile timezone)
-MORNING_DIGEST_ENABLED=true
-MORNING_DIGEST_HOUR=8
-MORNING_DIGEST_MINUTE=0
-
-# ComfyUI on GPU PC (Anima split-model — как в aiChatBot)
-COMFYUI_BASE_URL=http://192.168.1.109:8188
-COMFYUI_ENABLED=true
-# Anima: UNET+CLIP+VAE, CHECKPOINT пустой. Для SD1.5/Pony — задай CHECKPOINT, очисти UNET.
-COMFYUI_CHECKPOINT=
-COMFYUI_UNET=anima-preview3-base.safetensors
-COMFYUI_CLIP=qwen_3_06b_base.safetensors
-COMFYUI_VAE=qwen_image_vae.safetensors
-COMFYUI_STYLE_LORA=anima-preview-3-masterpieces-v5.safetensors
-COMFYUI_STYLE_LORA_WEIGHT=0.7
-COMFYUI_STEPS=30
-COMFYUI_CFG=4
-COMFYUI_SAMPLER=er_sde
-COMFYUI_SCHEDULER=simple
-COMFYUI_WIDTH=1024
-COMFYUI_HEIGHT=720
-COMFYUI_NEGATIVE_PROMPT=worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia
-COMFYUI_ROFL_ENABLED=true
-COMFYUI_ROFL_MAX_PER_DAY=1
-COMFYUI_ROFL_PROBABILITY=0.15
-COMFYUI_ROFL_MIN_INTERVAL_HOURS=12
-GENERATED_MEDIA_DIR=./data/generated
-
-# RP Chat (aiChatBot) — генерация картинок + sd-prompt; persona_id в карточке персонажа
-RP_CHAT_BASE_URL=http://host.docker.internal:8201
-RP_CHAT_ENABLED=true
-RP_CHAT_TIMEOUT_SEC=300
-
-# Netdata on server
-NETDATA_BASE_URL=http://host.docker.internal:19999
-NETDATA_PUBLIC_URL=
-NETDATA_ALERTS_ENABLED=true
-NETDATA_POLL_INTERVAL_SEC=120
-
-# Vector DB (phase 3)
-QDRANT_PORT=6333
-QDRANT_GRPC_PORT=6334
+# Server (internal bind inside containers)
+HOST=0.0.0.0
+BACKEND_INTERNAL_PORT=8080
+FRONTEND_INTERNAL_PORT=80
+
+# External ports on the host (docker compose publish)
+BACKEND_PORT=8080
+FRONTEND_PORT=3080
+VITE_DEV_PORT=5173
+
+# OpenRouter
+OPENROUTER_API_KEY=sk-or-v1-your-key-here
+OPENROUTER_MODEL=deepseek/deepseek-chat
+# deepseek/deepseek-v4-pro — сильная модель, tools поддерживаются:
+# OPENROUTER_MODEL=deepseek/deepseek-v4-pro
+OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
+OPENROUTER_TOOLS_ENABLED=true
+# none = без thinking (быстрее, стабильнее с tools). low|medium|high|xhigh — reasoning.
+OPENROUTER_REASONING_EFFORT=none
+# JSON-экстракция памяти отдельной моделью (если основная капризничает):
+# MEMORY_EXTRACT_MODEL=deepseek/deepseek-chat
+
+# App
+DATABASE_URL=sqlite:///./data/assistant.db
+CORS_ORIGINS=http://localhost:5173,http://localhost:8080,http://localhost:3080
+SYSTEM_PROMPT_PATH=./prompts/assistant.md
+MEMORY_AUTO_EXTRACT=true
+
+# Multi-user (API token auth)
+DEFAULT_USER_USERNAME=owner
+DEFAULT_USER_DISPLAY_NAME=
+DEFAULT_API_TOKEN=change-me-to-long-random-string
+AUTH_REQUIRED=true
+# Опционально для dev (автовход без /login). В prod оставьте пустым.
+VITE_API_TOKEN=
+
+# Fitness (wger + Open Food Facts — public HTTPS, no proxy)
+WGER_BASE_URL=https://wger.de/api/v2
+OPENFOODFACTS_BASE_URL=https://world.openfoodfacts.org
+FITNESS_REMINDERS_ENABLED=true
+REMINDERS_ENABLED=true
+
+# Taiga (on host :9000, nginx → taiga.grigowashere.ru)
+TAIGA_BASE_URL=http://host.docker.internal:9000
+TAIGA_USERNAME=your_taiga_user
+TAIGA_PASSWORD=your_taiga_password
+TAIGA_PUBLIC_URL=https://taiga.grigowashere.ru
+
+# Gitea (on host :3000, nginx → git.grigowashere.ru)
+GITEA_BASE_URL=http://host.docker.internal:3000
+GITEA_TOKEN=your_gitea_api_token
+GITEA_PUBLIC_URL=https://git.grigowashere.ru
+GITEA_WEBHOOK_SECRET=generate_a_random_secret
+
+# Gitea webhook URL (repo Settings → Webhooks):
+# https://assistant.your-domain/api/v1/webhooks/gitea   ← nginx → 127.0.0.1:BACKEND_PORT
+# http://172.17.0.1:8202/api/v1/webhooks/gitea        ← если Gitea в Docker (не 127.0.0.1!)
+
+REPOS_DIR=/data/repos
+
+# Homelab — GPU PC 192.168.1.109
+OPENMETEO_BASE_URL=http://192.168.1.109:8085
+WEATHER_LAT=59.9343
+WEATHER_LON=30.3351
+WEATHER_LOCATION_NAME=Санкт-Петербург
+WEATHER_CACHE_SEC=300
+
+# News RSS (comma-separated)
+NEWS_RSS_URLS=https://habr.com/ru/rss/all/all/,https://www.reddit.com/r/programming/.rss
+NEWS_CACHE_SEC=1800
+NEWS_MAX_ITEMS=7
+
+# Morning digest (Europe/Moscow or user profile timezone)
+MORNING_DIGEST_ENABLED=true
+MORNING_DIGEST_HOUR=8
+MORNING_DIGEST_MINUTE=0
+
+# ComfyUI on GPU PC (Anima split-model — как в aiChatBot)
+COMFYUI_BASE_URL=http://192.168.1.109:8188
+COMFYUI_ENABLED=true
+# Anima: UNET+CLIP+VAE, CHECKPOINT пустой. Для SD1.5/Pony — задай CHECKPOINT, очисти UNET.
+COMFYUI_CHECKPOINT=
+COMFYUI_UNET=anima-preview3-base.safetensors
+COMFYUI_CLIP=qwen_3_06b_base.safetensors
+COMFYUI_VAE=qwen_image_vae.safetensors
+COMFYUI_STYLE_LORA=anima-preview-3-masterpieces-v5.safetensors
+COMFYUI_STYLE_LORA_WEIGHT=0.7
+COMFYUI_STEPS=30
+COMFYUI_CFG=4
+COMFYUI_SAMPLER=er_sde
+COMFYUI_SCHEDULER=simple
+COMFYUI_WIDTH=1024
+COMFYUI_HEIGHT=720
+COMFYUI_NEGATIVE_PROMPT=worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia
+COMFYUI_ROFL_ENABLED=true
+COMFYUI_ROFL_MAX_PER_DAY=1
+COMFYUI_ROFL_PROBABILITY=0.15
+COMFYUI_ROFL_MIN_INTERVAL_HOURS=12
+GENERATED_MEDIA_DIR=./data/generated
+
+# RP Chat (aiChatBot) — генерация картинок + sd-prompt; persona_id в карточке персонажа
+RP_CHAT_BASE_URL=http://host.docker.internal:8201
+RP_CHAT_ENABLED=true
+RP_CHAT_TIMEOUT_SEC=300
+
+# Netdata on server
+NETDATA_BASE_URL=http://host.docker.internal:19999
+NETDATA_PUBLIC_URL=
+NETDATA_ALERTS_ENABLED=true
+NETDATA_POLL_INTERVAL_SEC=120
+
+# Vector DB (phase 3)
+QDRANT_PORT=6333
+QDRANT_GRPC_PORT=6334
+
+# RAG / embeddings
+QDRANT_URL=http://qdrant:6333
+EMBEDDING_MODEL=openai/text-embedding-3-small
+RAG_ENABLED=true
+RAG_TOP_K=8
+MEMORY_FACTS_IN_CONTEXT=8

.env.example.refactor_bak

@@ -0,0 +1,106 @@
+# Server (internal bind inside containers)
+HOST=0.0.0.0
+BACKEND_INTERNAL_PORT=8080
+FRONTEND_INTERNAL_PORT=80
+
+# External ports on the host (docker compose publish)
+BACKEND_PORT=8080
+FRONTEND_PORT=3080
+VITE_DEV_PORT=5173
+
+# OpenRouter
+OPENROUTER_API_KEY=sk-or-v1-your-key-here
+OPENROUTER_MODEL=deepseek/deepseek-chat
+# deepseek/deepseek-v4-pro — сильная модель, tools поддерживаются:
+# OPENROUTER_MODEL=deepseek/deepseek-v4-pro
+OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
+OPENROUTER_TOOLS_ENABLED=true
+# none = без thinking (быстрее, стабильнее с tools). low|medium|high|xhigh — reasoning.
+OPENROUTER_REASONING_EFFORT=none
+# JSON-экстракция памяти отдельной моделью (если основная капризничает):
+# MEMORY_EXTRACT_MODEL=deepseek/deepseek-chat
+
+# App
+DATABASE_URL=sqlite:///./data/assistant.db
+CORS_ORIGINS=http://localhost:5173,http://localhost:8080,http://localhost:3080
+SYSTEM_PROMPT_PATH=./prompts/assistant.md
+MEMORY_AUTO_EXTRACT=true
+
+# Fitness (wger + Open Food Facts — public HTTPS, no proxy)
+WGER_BASE_URL=https://wger.de/api/v2
+OPENFOODFACTS_BASE_URL=https://world.openfoodfacts.org
+FITNESS_REMINDERS_ENABLED=true
+REMINDERS_ENABLED=true
+
+# Taiga (on host :9000, nginx → taiga.grigowashere.ru)
+TAIGA_BASE_URL=http://host.docker.internal:9000
+TAIGA_USERNAME=your_taiga_user
+TAIGA_PASSWORD=your_taiga_password
+TAIGA_PUBLIC_URL=https://taiga.grigowashere.ru
+
+# Gitea (on host :3000, nginx → git.grigowashere.ru)
+GITEA_BASE_URL=http://host.docker.internal:3000
+GITEA_TOKEN=your_gitea_api_token
+GITEA_PUBLIC_URL=https://git.grigowashere.ru
+GITEA_WEBHOOK_SECRET=generate_a_random_secret
+
+# Gitea webhook URL (repo Settings → Webhooks):
+# https://assistant.your-domain/api/v1/webhooks/gitea   ← nginx → 127.0.0.1:BACKEND_PORT
+# http://172.17.0.1:8202/api/v1/webhooks/gitea        ← если Gitea в Docker (не 127.0.0.1!)
+
+REPOS_DIR=/data/repos
+
+# Homelab — GPU PC 192.168.1.109
+OPENMETEO_BASE_URL=http://192.168.1.109:8085
+WEATHER_LAT=59.9343
+WEATHER_LON=30.3351
+WEATHER_LOCATION_NAME=Санкт-Петербург
+WEATHER_CACHE_SEC=300
+
+# News RSS (comma-separated)
+NEWS_RSS_URLS=https://habr.com/ru/rss/all/all/,https://www.reddit.com/r/programming/.rss
+NEWS_CACHE_SEC=1800
+NEWS_MAX_ITEMS=7
+
+# Morning digest (Europe/Moscow or user profile timezone)
+MORNING_DIGEST_ENABLED=true
+MORNING_DIGEST_HOUR=8
+MORNING_DIGEST_MINUTE=0
+
+# ComfyUI on GPU PC (Anima split-model — как в aiChatBot)
+COMFYUI_BASE_URL=http://192.168.1.109:8188
+COMFYUI_ENABLED=true
+# Anima: UNET+CLIP+VAE, CHECKPOINT пустой. Для SD1.5/Pony — задай CHECKPOINT, очисти UNET.
+COMFYUI_CHECKPOINT=
+COMFYUI_UNET=anima-preview3-base.safetensors
+COMFYUI_CLIP=qwen_3_06b_base.safetensors
+COMFYUI_VAE=qwen_image_vae.safetensors
+COMFYUI_STYLE_LORA=anima-preview-3-masterpieces-v5.safetensors
+COMFYUI_STYLE_LORA_WEIGHT=0.7
+COMFYUI_STEPS=30
+COMFYUI_CFG=4
+COMFYUI_SAMPLER=er_sde
+COMFYUI_SCHEDULER=simple
+COMFYUI_WIDTH=1024
+COMFYUI_HEIGHT=720
+COMFYUI_NEGATIVE_PROMPT=worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia
+COMFYUI_ROFL_ENABLED=true
+COMFYUI_ROFL_MAX_PER_DAY=1
+COMFYUI_ROFL_PROBABILITY=0.15
+COMFYUI_ROFL_MIN_INTERVAL_HOURS=12
+GENERATED_MEDIA_DIR=./data/generated
+
+# RP Chat (aiChatBot) — генерация картинок + sd-prompt; persona_id в карточке персонажа
+RP_CHAT_BASE_URL=http://host.docker.internal:8201
+RP_CHAT_ENABLED=true
+RP_CHAT_TIMEOUT_SEC=300
+
+# Netdata on server
+NETDATA_BASE_URL=http://host.docker.internal:19999
+NETDATA_PUBLIC_URL=
+NETDATA_ALERTS_ENABLED=true
+NETDATA_POLL_INTERVAL_SEC=120
+
+# Vector DB (phase 3)
+QDRANT_PORT=6333
+QDRANT_GRPC_PORT=6334

backend/_p1.py

@@ -0,0 +1,34 @@
+import pathlib
+ROOT = pathlib.Path(".").resolve()
+svc = ROOT / "app/fitness/service.py"
+text = svc.read_text(encoding="utf-8")
+old = """from app.db.models import (
+    BodyMetric,
+    FitnessProfile,
+    FitnessReminder,
+    FoodLog,
+    WaterLog,
+    WorkoutLog,
+)
+from app.fitness.calculators import compute_targets, one_rep_max"""
+new = """from app.db.models import (
+    BodyMetric,
+    FitnessProfile,
+    FitnessReminder,
+    FoodLog,
+    StepLog,
+    WaterLog,
+    WorkoutLog,
+)
+from app.fitness.activity_budget import (
+    build_base_targets,
+    compute_activity_bonus,
+    estimate_workout_active_kcal,
+    scale_targets,
+)
+from app.fitness.calculators import compute_targets, one_rep_max"""
+if old not in text:
+    raise SystemExit("import block missing")
+text = text.replace(old, new, 1)
+svc.write_text(text, encoding="utf-8")
+print("ok imports")

backend/app/api/chat_schemas.py

@@ -0,0 +1,12 @@
+from pydantic import BaseModel
+
+from app.api.schemas import MessageOut
+
+
+class MessagesPageOut(BaseModel):
+    messages: list[MessageOut]
+    has_more: bool
+
+
+class GenerationStatusOut(BaseModel):
+    active: bool

backend/app/api/routes/__init__.py

@@ -1,17 +1,20 @@
-from fastapi import APIRouter
-
-from app.api.routes import character, chat, fitness, health, homelab, media, memory, pomodoro, projects, reminders, shopping, webhooks
-
-api_router = APIRouter(prefix="/api/v1")
-api_router.include_router(health.router, tags=["health"])
-api_router.include_router(homelab.router, tags=["homelab"])
-api_router.include_router(chat.router, prefix="/chat", tags=["chat"])
-api_router.include_router(pomodoro.router, prefix="/pomodoro", tags=["pomodoro"])
-api_router.include_router(character.router, tags=["character"])
-api_router.include_router(projects.router, tags=["projects"])
-api_router.include_router(memory.router, tags=["memory"])
-api_router.include_router(fitness.router, tags=["fitness"])
-api_router.include_router(shopping.router, prefix="/shopping", tags=["shopping"])
-api_router.include_router(reminders.router, prefix="/reminders", tags=["reminders"])
-api_router.include_router(webhooks.router, tags=["webhooks"])
-api_router.include_router(media.router, tags=["media"])
+from fastapi import APIRouter
+
+from app.api.routes import auth, character, chat, documents, fitness, health, homelab, media, memory, pomodoro, projects, reminders, settings, shopping, webhooks
+
+api_router = APIRouter(prefix="/api/v1")
+api_router.include_router(health.router, tags=["health"])
+api_router.include_router(auth.router)
+api_router.include_router(homelab.router, tags=["homelab"])
+api_router.include_router(chat.router, prefix="/chat", tags=["chat"])
+api_router.include_router(pomodoro.router, prefix="/pomodoro", tags=["pomodoro"])
+api_router.include_router(character.router, tags=["character"])
+api_router.include_router(projects.router, tags=["projects"])
+api_router.include_router(memory.router, tags=["memory"])
+api_router.include_router(fitness.router, tags=["fitness"])
+api_router.include_router(shopping.router, prefix="/shopping", tags=["shopping"])
+api_router.include_router(reminders.router, prefix="/reminders", tags=["reminders"])
+api_router.include_router(webhooks.router, tags=["webhooks"])
+api_router.include_router(media.router, tags=["media"])
+api_router.include_router(settings.router, tags=["settings"])
+api_router.include_router(documents.router, tags=["documents"])

backend/app/api/routes/__init__.py.refactor_bak

@@ -0,0 +1,17 @@
+from fastapi import APIRouter
+
+from app.api.routes import character, chat, fitness, health, homelab, media, memory, pomodoro, projects, reminders, shopping, webhooks
+
+api_router = APIRouter(prefix="/api/v1")
+api_router.include_router(health.router, tags=["health"])
+api_router.include_router(homelab.router, tags=["homelab"])
+api_router.include_router(chat.router, prefix="/chat", tags=["chat"])
+api_router.include_router(pomodoro.router, prefix="/pomodoro", tags=["pomodoro"])
+api_router.include_router(character.router, tags=["character"])
+api_router.include_router(projects.router, tags=["projects"])
+api_router.include_router(memory.router, tags=["memory"])
+api_router.include_router(fitness.router, tags=["fitness"])
+api_router.include_router(shopping.router, prefix="/shopping", tags=["shopping"])
+api_router.include_router(reminders.router, prefix="/reminders", tags=["reminders"])
+api_router.include_router(webhooks.router, tags=["webhooks"])
+api_router.include_router(media.router, tags=["media"])

backend/app/api/routes/auth.py

@@ -0,0 +1,73 @@
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel, Field
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.auth.service import create_user, find_user_by_token, user_to_dict
+from app.db.base import get_db
+from app.db.models import User
+
+router = APIRouter(prefix="/auth", tags=["auth"])
+
+
+class LoginRequest(BaseModel):
+    token: str = Field(min_length=8, max_length=256)
+
+
+class CreateUserRequest(BaseModel):
+    username: str = Field(min_length=2, max_length=64)
+    display_name: str = ""
+    token: str | None = Field(default=None, min_length=8, max_length=256)
+
+
+@router.post("/login")
+def login(payload: LoginRequest, db: Session = Depends(get_db)) -> dict[str, Any]:
+    user = find_user_by_token(db, payload.token)
+    if not user:
+        raise HTTPException(status_code=401, detail="Неверный токен")
+    return {"ok": True, "user": user_to_dict(user), "token": payload.token.strip()}
+
+
+@router.get("/me")
+def me(user: User = Depends(get_current_user)) -> dict[str, Any]:
+    return {"ok": True, "user": user_to_dict(user)}
+
+
+@router.get("/users")
+def list_users(
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    rows = db.scalars(select(User).where(User.is_active.is_(True)).order_by(User.id)).all()
+    return {
+        "ok": True,
+        "users": [user_to_dict(row) for row in rows],
+        "current_user_id": user.id,
+    }
+
+
+@router.post("/users")
+def register_user(
+    payload: CreateUserRequest,
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        new_user, plain_token = create_user(
+            db,
+            username=payload.username,
+            display_name=payload.display_name,
+            api_token=payload.token,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+    return {
+        "ok": True,
+        "user": user_to_dict(new_user),
+        "token": plain_token,
+        "created_by": user.username,
+    }

backend/app/api/routes/character.py

@@ -1,62 +1,80 @@
-from typing import Any
-
-from fastapi import APIRouter, HTTPException
-from pydantic import BaseModel, Field
-
-from app.character.service import CharacterService
-
-router = APIRouter()
-
-
-class CharacterCardData(BaseModel):
-    name: str = "Ассистент"
-    description: str = ""
-    personality: str = ""
-    scenario: str = ""
-    first_mes: str = ""
-    mes_example: str = ""
-    system_prompt: str = ""
-    post_history_instructions: str = ""
-    tags: list[str] = Field(default_factory=list)
-    creator: str = ""
-    creator_notes: str = ""
-    alternate_greetings: list[str] = Field(default_factory=list)
-    character_version: str = "1.0"
-    appearance_tags: str = ""
-    appearance_prose: str = ""
-    lora_name: str = ""
-    lora_weight: float = 0.8
-    rp_persona_id: str = ""
-    sd_enabled: bool = True
-
-
-class CharacterCardV2(BaseModel):
-    spec: str = "chara_card_v2"
-    spec_version: str = "2.0"
-    data: CharacterCardData
-
-
-@router.get("/character")
-def get_character() -> dict[str, Any]:
-    return CharacterService().get_card()
-
-
-@router.put("/character")
-def update_character(payload: CharacterCardV2) -> dict[str, Any]:
-    return CharacterService().save_card(payload.model_dump())
-
-
-@router.get("/character/prompt")
-def get_character_prompt() -> dict[str, str]:
-    service = CharacterService()
-    return {
-        "system_prompt": service.get_system_prompt(),
-        "first_mes": service.get_card().get("data", {}).get("first_mes", ""),
-    }
-
-
-@router.post("/character/import")
-def import_character(payload: dict[str, Any]) -> dict[str, Any]:
-    if not payload:
-        raise HTTPException(status_code=400, detail="Empty card")
-    return CharacterService().save_card(payload)
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.character.service import CharacterService
+from app.db.base import get_db
+from app.db.models import User
+
+router = APIRouter()
+
+
+class CharacterCardData(BaseModel):
+    name: str = "Ассистент"
+    description: str = ""
+    personality: str = ""
+    scenario: str = ""
+    first_mes: str = ""
+    mes_example: str = ""
+    system_prompt: str = ""
+    post_history_instructions: str = ""
+    tags: list[str] = Field(default_factory=list)
+    creator: str = ""
+    creator_notes: str = ""
+    alternate_greetings: list[str] = Field(default_factory=list)
+    character_version: str = "1.0"
+    appearance_tags: str = ""
+    appearance_prose: str = ""
+    lora_name: str = ""
+    lora_weight: float = 0.8
+    rp_persona_id: str = ""
+    sd_enabled: bool = True
+
+
+class CharacterCardV2(BaseModel):
+    spec: str = "chara_card_v2"
+    spec_version: str = "2.0"
+    data: CharacterCardData
+
+
+@router.get("/character")
+def get_character(
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return CharacterService(db, user.id).get_card()
+
+
+@router.put("/character")
+def update_character(
+    payload: CharacterCardV2,
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return CharacterService(db, user.id).save_card(payload.model_dump())
+
+
+@router.get("/character/prompt")
+def get_character_prompt(
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, str]:
+    service = CharacterService(db, user.id)
+    return {
+        "system_prompt": service.get_system_prompt(),
+        "first_mes": service.get_card().get("data", {}).get("first_mes", ""),
+    }
+
+
+@router.post("/character/import")
+def import_character(
+    payload: dict[str, Any],
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    if not payload:
+        raise HTTPException(status_code=400, detail="Empty card")
+    return CharacterService(db, user.id).save_card(payload)

backend/app/api/routes/chat.py

@@ -1,70 +1,158 @@
-from fastapi import APIRouter, Depends, HTTPException
-from fastapi.responses import StreamingResponse
-from sqlalchemy.orm import Session
-
-from app.api.schemas import MessageCreate, MessageOut, SessionCreate, SessionDetailOut, SessionOut
-from app.chat.service import ChatService
-from app.db.base import get_db
-
-router = APIRouter()
-
-
-@router.post("/sessions", response_model=SessionOut)
-def create_session(payload: SessionCreate, db: Session = Depends(get_db)) -> SessionOut:
-    service = ChatService(db)
-    return service.create_session(title=payload.title)
-
-
-@router.get("/sessions", response_model=list[SessionOut])
-def list_sessions(db: Session = Depends(get_db)) -> list[SessionOut]:
-    service = ChatService(db)
-    return service.list_sessions()
-
-
-@router.get("/sessions/{session_id}", response_model=SessionDetailOut)
-def get_session(session_id: int, db: Session = Depends(get_db)) -> SessionDetailOut:
-    service = ChatService(db)
-    session = service.get_session(session_id)
-    if not session:
-        raise HTTPException(status_code=404, detail="Session not found")
-    return session
-
-
-@router.delete("/sessions/{session_id}")
-def delete_session(session_id: int, db: Session = Depends(get_db)) -> dict[str, bool]:
-    service = ChatService(db)
-    if not service.delete_session(session_id):
-        raise HTTPException(status_code=404, detail="Session not found")
-    return {"ok": True}
-
-
-@router.post("/sessions/{session_id}/messages")
-async def send_message(
-    session_id: int,
-    payload: MessageCreate,
-    db: Session = Depends(get_db),
-) -> StreamingResponse:
-    service = ChatService(db)
-    if not service.get_session(session_id):
-        raise HTTPException(status_code=404, detail="Session not found")
-
-    # Сохраняем user до стрима: иначе при обрыве SSE сообщение не попадает в БД.
-    service.save_user_message(session_id, payload.content)
-
-    async def event_stream():
-        async for chunk in service.stream_response(
-            session_id,
-            payload.content,
-            user_message_saved=True,
-        ):
-            yield chunk
-
-    return StreamingResponse(
-        event_stream(),
-        media_type="text/event-stream",
-        headers={
-            "Cache-Control": "no-cache",
-            "Connection": "keep-alive",
-            "X-Accel-Buffering": "no",
-        },
-    )
+import asyncio
+
+from fastapi import APIRouter, Depends, HTTPException
+from fastapi.responses import StreamingResponse
+from sqlalchemy.orm import Session
+
+from app.api.chat_schemas import GenerationStatusOut, MessagesPageOut
+from app.api.schemas import (
+    MessageCreate,
+    SessionCreate,
+    SessionDetailOut,
+    SessionOut,
+)
+from app.chat.generation import (
+    GenerationBusyError,
+    get_active_handle,
+    is_generation_active,
+    start_generation,
+    subscribe_generation,
+)
+from app.chat.service import ChatService
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+
+router = APIRouter()
+
+
+@router.post("/sessions", response_model=SessionOut)
+def create_session(payload: SessionCreate, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> SessionOut:
+    service = ChatService(db, user.id)
+    return service.create_session(title=payload.title)
+
+
+@router.get("/sessions", response_model=list[SessionOut])
+def list_sessions(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[SessionOut]:
+    service = ChatService(db, user.id)
+    return service.list_sessions()
+
+
+@router.get("/sessions/{session_id}", response_model=SessionDetailOut)
+def get_session(session_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> SessionDetailOut:
+    service = ChatService(db, user.id)
+    session = service.get_session(session_id)
+    if not session:
+        raise HTTPException(status_code=404, detail="Session not found")
+    return session
+
+
+@router.get("/sessions/{session_id}/messages", response_model=MessagesPageOut)
+def list_messages(
+    session_id: int,
+    limit: int = 30,
+    before_id: int | None = None,
+    after_id: int | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> MessagesPageOut:
+    service = ChatService(db, user.id)
+    if not service.get_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+    messages, has_more = service.list_messages(
+        session_id,
+        limit=min(max(limit, 1), 100),
+        before_id=before_id,
+        after_id=after_id,
+    )
+    return MessagesPageOut(messages=messages, has_more=has_more)
+
+
+@router.get("/sessions/{session_id}/generation", response_model=GenerationStatusOut)
+def generation_status(session_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> GenerationStatusOut:
+    service = ChatService(db, user.id)
+    if not service.get_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+    return GenerationStatusOut(active=is_generation_active(session_id))
+
+
+@router.get("/sessions/{session_id}/generation/stream")
+async def generation_stream(session_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> StreamingResponse:
+    service = ChatService(db, user.id)
+    if not service.get_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+
+    handle = get_active_handle(session_id)
+    if not handle:
+        raise HTTPException(status_code=404, detail="No active generation")
+
+    async def event_stream():
+        async for chunk in subscribe_generation(handle):
+            yield chunk
+
+    return StreamingResponse(
+        event_stream(),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+        },
+    )
+
+
+@router.delete("/sessions/{session_id}")
+def delete_session(session_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, bool]:
+    service = ChatService(db, user.id)
+    if not service.delete_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+    return {"ok": True}
+
+
+@router.post("/sessions/{session_id}/messages")
+async def send_message(
+    session_id: int,
+    payload: MessageCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> StreamingResponse:
+    service = ChatService(db, user.id)
+    if not service.get_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+
+    if is_generation_active(session_id):
+        raise HTTPException(status_code=409, detail="Generation already in progress")
+
+    # Сохраняем user до стрима: иначе при обрыве SSE сообщение не попадает в БД.
+    service.save_user_message(session_id, payload.content)
+
+    try:
+        handle = await start_generation(session_id, user.id, payload.content)
+    except GenerationBusyError:
+        raise HTTPException(status_code=409, detail="Generation already in progress") from None
+
+    async def event_stream():
+        try:
+            async for chunk in subscribe_generation(handle):
+                yield chunk
+        except asyncio.CancelledError:
+            raise
+
+    return StreamingResponse(
+        event_stream(),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+        },
+    )
+
+
+@router.get("/sessions/{session_id}/context-preview")
+def context_preview(
+    session_id: int,
+    query: str | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict:
+    service = ChatService(db, user.id)
+    return service.context_preview(session_id, query=query)
+

backend/app/api/routes/chat.py.refactor_bak

@@ -0,0 +1,70 @@
+from fastapi import APIRouter, Depends, HTTPException
+from fastapi.responses import StreamingResponse
+from sqlalchemy.orm import Session
+
+from app.api.schemas import MessageCreate, MessageOut, SessionCreate, SessionDetailOut, SessionOut
+from app.chat.service import ChatService
+from app.db.base import get_db
+
+router = APIRouter()
+
+
+@router.post("/sessions", response_model=SessionOut)
+def create_session(payload: SessionCreate, db: Session = Depends(get_db)) -> SessionOut:
+    service = ChatService(db)
+    return service.create_session(title=payload.title)
+
+
+@router.get("/sessions", response_model=list[SessionOut])
+def list_sessions(db: Session = Depends(get_db)) -> list[SessionOut]:
+    service = ChatService(db)
+    return service.list_sessions()
+
+
+@router.get("/sessions/{session_id}", response_model=SessionDetailOut)
+def get_session(session_id: int, db: Session = Depends(get_db)) -> SessionDetailOut:
+    service = ChatService(db)
+    session = service.get_session(session_id)
+    if not session:
+        raise HTTPException(status_code=404, detail="Session not found")
+    return session
+
+
+@router.delete("/sessions/{session_id}")
+def delete_session(session_id: int, db: Session = Depends(get_db)) -> dict[str, bool]:
+    service = ChatService(db)
+    if not service.delete_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+    return {"ok": True}
+
+
+@router.post("/sessions/{session_id}/messages")
+async def send_message(
+    session_id: int,
+    payload: MessageCreate,
+    db: Session = Depends(get_db),
+) -> StreamingResponse:
+    service = ChatService(db)
+    if not service.get_session(session_id):
+        raise HTTPException(status_code=404, detail="Session not found")
+
+    # Сохраняем user до стрима: иначе при обрыве SSE сообщение не попадает в БД.
+    service.save_user_message(session_id, payload.content)
+
+    async def event_stream():
+        async for chunk in service.stream_response(
+            session_id,
+            payload.content,
+            user_message_saved=True,
+        ):
+            yield chunk
+
+    return StreamingResponse(
+        event_stream(),
+        media_type="text/event-stream",
+        headers={
+            "Cache-Control": "no-cache",
+            "Connection": "keep-alive",
+            "X-Accel-Buffering": "no",
+        },
+    )

backend/app/api/routes/documents.py

@@ -0,0 +1,51 @@
+from typing import Any
+
+from fastapi import APIRouter, Depends, File, Form, HTTPException, UploadFile
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.db.models import Document
+from app.rag.ingest import ingest_document_file
+
+router = APIRouter()
+
+
+@router.get("/documents")
+def list_documents(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[dict[str, Any]]:
+    docs = db.scalars(select(Document).where(Document.user_id == user.id).order_by(Document.created_at.desc())).all()
+    return [
+        {
+            "id": d.id,
+            "title": d.title,
+            "filename": d.filename,
+            "size_bytes": d.size_bytes,
+            "created_at": d.created_at.isoformat() if d.created_at else None,
+        }
+        for d in docs
+    ]
+
+
+@router.post("/documents/upload")
+async def upload_document(
+    file: UploadFile = File(...),
+    title: str = Form(""),
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    raw = await file.read()
+    if not raw:
+        raise HTTPException(status_code=400, detail="Empty file")
+    try:
+        doc = await ingest_document_file(
+            db,
+            user_id=user.id,
+            title=title.strip() or (file.filename or "document"),
+            filename=file.filename or "upload.txt",
+            raw_bytes=raw,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+    return {"ok": True, "document": doc}

backend/app/api/routes/fitness.py

@@ -1,223 +1,308 @@
-from datetime import date
-from typing import Any
-
-from fastapi import APIRouter, Depends, HTTPException
-from pydantic import BaseModel, Field
-from sqlalchemy.orm import Session
-
-from app.db.base import get_db
-from app.fitness.service import FitnessService
-from app.fitness.structuring import structure_meal, structure_workout
-from app.integrations.openfoodfacts import OpenFoodFactsClient
-from app.integrations.wger import WgerClient
-
-router = APIRouter()
-
-
-class ProfileUpdate(BaseModel):
-    sex: str | None = None
-    age: int | None = None
-    height_cm: float | None = None
-    weight_kg: float | None = None
-    activity_level: str | None = None
-    goal: str | None = None
-    target_weight_kg: float | None = None
-    weekly_workouts: int | None = None
-
-
-class MealCreate(BaseModel):
-    text: str = Field(min_length=1)
-    meal_type: str | None = None
-
-
-class WaterCreate(BaseModel):
-    amount_ml: int = Field(gt=0)
-
-
-class WeightCreate(BaseModel):
-    weight_kg: float = Field(gt=0)
-    body_fat_pct: float | None = None
-    chest_cm: float | None = None
-    waist_cm: float | None = None
-    notes: str = ""
-
-
-class WorkoutCreate(BaseModel):
-    text: str = Field(min_length=1)
-
-
-class ReminderUpdate(BaseModel):
-    enabled: bool | None = None
-    hour: int | None = Field(default=None, ge=0, le=23)
-    minute: int | None = Field(default=None, ge=0, le=59)
-    interval_hours: int | None = Field(default=None, ge=1, le=12)
-
-
-@router.get("/fitness")
-def get_snapshot(db: Session = Depends(get_db)) -> dict[str, Any]:
-    return FitnessService(db).snapshot()
-
-
-@router.get("/fitness/summary")
-def get_summary(
-    day: str | None = None,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    d = date.fromisoformat(day) if day else None
-    return FitnessService(db).get_daily_summary(d)
-
-
-@router.get("/fitness/history")
-def get_history(
-    days: int = 7,
-    end: str | None = None,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    end_day = date.fromisoformat(end) if end else None
-    return FitnessService(db).get_history(days=days, end_day=end_day)
-
-
-@router.get("/fitness/profile")
-def get_profile(db: Session = Depends(get_db)) -> dict[str, Any]:
-    profile = FitnessService(db).get_profile()
-    return profile or {"configured": False}
-
-
-@router.put("/fitness/profile")
-def update_profile(
-    payload: ProfileUpdate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    return FitnessService(db).set_profile(payload.model_dump(exclude_none=True))
-
-
-@router.post("/fitness/profile/calc")
-def calc_targets(
-    payload: ProfileUpdate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    params = payload.model_dump(exclude_none=True)
-    if not params:
-        raise HTTPException(status_code=400, detail="No parameters")
-    return FitnessService(db).calc_targets(params)
-
-
-@router.post("/fitness/meals")
-async def create_meal(
-    payload: MealCreate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    service = FitnessService(db)
-    try:
-        structured = await structure_meal(payload.text)
-    except Exception as exc:
-        raise HTTPException(status_code=502, detail=str(exc)) from exc
-    return service.log_meal(
-        description=structured.get("description") or payload.text,
-        meal_type=payload.meal_type or structured.get("meal_type") or "snack",
-        calories=float(structured.get("calories") or 0),
-        protein_g=float(structured.get("protein_g") or 0),
-        fat_g=float(structured.get("fat_g") or 0),
-        carbs_g=float(structured.get("carbs_g") or 0),
-        source="llm",
-        estimated=bool(structured.get("estimated", True)),
-    )
-
-
-@router.post("/fitness/water")
-def create_water(
-    payload: WaterCreate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    return FitnessService(db).log_water(payload.amount_ml)
-
-
-@router.post("/fitness/weight")
-def create_weight(
-    payload: WeightCreate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    return FitnessService(db).log_weight(
-        payload.weight_kg,
-        body_fat_pct=payload.body_fat_pct,
-        chest_cm=payload.chest_cm,
-        waist_cm=payload.waist_cm,
-        notes=payload.notes,
-    )
-
-
-@router.post("/fitness/workouts")
-async def create_workout(
-    payload: WorkoutCreate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    service = FitnessService(db)
-    try:
-        structured = await structure_workout(payload.text)
-    except Exception as exc:
-        raise HTTPException(status_code=502, detail=str(exc)) from exc
-    return service.log_workout(
-        title=structured.get("title") or "Тренировка",
-        notes=structured.get("notes") or payload.text,
-        duration_min=structured.get("duration_min"),
-        exercises=structured.get("exercises"),
-    )
-
-
-@router.get("/fitness/body-metrics")
-def list_metrics(
-    limit: int = 30,
-    db: Session = Depends(get_db),
-) -> list[dict[str, Any]]:
-    return FitnessService(db).list_body_metrics(limit=limit)
-
-
-@router.delete("/fitness/meals/{log_id}")
-def delete_meal(log_id: int, db: Session = Depends(get_db)) -> dict[str, bool]:
-    if not FitnessService(db).delete_food_log(log_id):
-        raise HTTPException(status_code=404, detail="Not found")
-    return {"ok": True}
-
-
-@router.delete("/fitness/water/{log_id}")
-def delete_water(log_id: int, db: Session = Depends(get_db)) -> dict[str, bool]:
-    if not FitnessService(db).delete_water_log(log_id):
-        raise HTTPException(status_code=404, detail="Not found")
-    return {"ok": True}
-
-
-@router.delete("/fitness/workouts/{log_id}")
-def delete_workout(log_id: int, db: Session = Depends(get_db)) -> dict[str, bool]:
-    if not FitnessService(db).delete_workout_log(log_id):
-        raise HTTPException(status_code=404, detail="Not found")
-    return {"ok": True}
-
-
-@router.get("/fitness/reminders")
-def list_reminders(db: Session = Depends(get_db)) -> list[dict[str, Any]]:
-    return FitnessService(db).list_reminders()
-
-
-@router.put("/fitness/reminders/{kind}")
-def update_reminder(
-    kind: str,
-    payload: ReminderUpdate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    return FitnessService(db).set_reminder(
-        kind,
-        enabled=payload.enabled,
-        hour=payload.hour,
-        minute=payload.minute,
-        interval_hours=payload.interval_hours,
-    )
-
-
-@router.get("/fitness/lookup/food")
-def lookup_food(q: str, limit: int = 5) -> list[dict[str, Any]]:
-    return OpenFoodFactsClient().search(q, limit=limit)
-
-
-@router.get("/fitness/lookup/exercise")
-def lookup_exercise(q: str, limit: int = 8) -> list[dict[str, Any]]:
-    return WgerClient().search_exercises(q, limit=limit)
+from datetime import date
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.fitness.service import FitnessService
+from app.fitness.structuring import structure_meal, structure_workout
+from app.integrations.openfoodfacts import OpenFoodFactsClient
+from app.integrations.wger import WgerClient
+
+router = APIRouter()
+
+
+class ProfileUpdate(BaseModel):
+    sex: str | None = None
+    age: int | None = None
+    height_cm: float | None = None
+    weight_kg: float | None = None
+    activity_level: str | None = None
+    goal: str | None = None
+    target_weight_kg: float | None = None
+    weekly_workouts: int | None = None
+    baseline_steps: int | None = None
+    baseline_workout_kcal: float | None = None
+
+
+class MealCreate(BaseModel):
+    text: str = Field(min_length=1)
+    meal_type: str | None = None
+
+
+class WaterCreate(BaseModel):
+    amount_ml: int = Field(gt=0)
+
+
+class WeightCreate(BaseModel):
+    weight_kg: float = Field(gt=0)
+    body_fat_pct: float | None = None
+    chest_cm: float | None = None
+    waist_cm: float | None = None
+    neck_cm: float | None = None
+    hip_cm: float | None = None
+    notes: str = ""
+    day: str | None = None
+    days_ago: int | None = Field(default=None, ge=0, le=90)
+    recorded_at: str | None = None
+
+
+class BodyCompositionCalc(BaseModel):
+    weight_kg: float | None = None
+    height_cm: float | None = None
+    sex: str | None = None
+    neck_cm: float | None = None
+    waist_cm: float | None = None
+    hip_cm: float | None = None
+    body_fat_pct: float | None = None
+
+
+class StepsCreate(BaseModel):
+    steps: int = Field(ge=0)
+    active_calories: float | None = None
+    notes: str = ""
+    day: str | None = None
+    days_ago: int | None = Field(default=None, ge=0, le=90)
+    logged_at: str | None = None
+
+
+class WorkoutCreate(BaseModel):
+    text: str = Field(min_length=1)
+    day: str | None = None
+    days_ago: int | None = Field(default=None, ge=0, le=90)
+    logged_at: str | None = None
+
+
+class ReminderUpdate(BaseModel):
+    enabled: bool | None = None
+    hour: int | None = Field(default=None, ge=0, le=23)
+    minute: int | None = Field(default=None, ge=0, le=59)
+    interval_hours: int | None = Field(default=None, ge=1, le=12)
+
+
+@router.get("/fitness")
+def get_snapshot(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    return FitnessService(db, user.id).snapshot()
+
+
+@router.get("/fitness/summary")
+def get_summary(
+    day: str | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    d = date.fromisoformat(day) if day else None
+    return FitnessService(db, user.id).get_daily_summary(d)
+
+
+@router.get("/fitness/workout-stats")
+def get_workout_stats(
+    days: int = 7,
+    end: str | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    end_day = date.fromisoformat(end) if end else None
+    return FitnessService(db, user.id).get_workout_stats(days=days, end_day=end_day)
+
+
+@router.get("/fitness/history")
+def get_history(
+    days: int = 7,
+    end: str | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    end_day = date.fromisoformat(end) if end else None
+    return FitnessService(db, user.id).get_history(days=days, end_day=end_day)
+
+
+@router.get("/fitness/profile")
+def get_profile(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    profile = FitnessService(db, user.id).get_profile()
+    return profile or {"configured": False}
+
+
+@router.put("/fitness/profile")
+def update_profile(
+    payload: ProfileUpdate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return FitnessService(db, user.id).set_profile(payload.model_dump(exclude_none=True))
+
+
+@router.post("/fitness/profile/calc")
+def calc_targets(
+    payload: ProfileUpdate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    params = payload.model_dump(exclude_none=True)
+    if not params:
+        raise HTTPException(status_code=400, detail="No parameters")
+    return FitnessService(db, user.id).calc_targets(params)
+
+
+@router.post("/fitness/meals")
+async def create_meal(
+    payload: MealCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    service = FitnessService(db, user.id)
+    try:
+        structured = await structure_meal(payload.text)
+    except Exception as exc:
+        raise HTTPException(status_code=502, detail=str(exc)) from exc
+    return service.log_meal(
+        description=structured.get("description") or payload.text,
+        meal_type=payload.meal_type or structured.get("meal_type") or "snack",
+        calories=float(structured.get("calories") or 0),
+        protein_g=float(structured.get("protein_g") or 0),
+        fat_g=float(structured.get("fat_g") or 0),
+        carbs_g=float(structured.get("carbs_g") or 0),
+        source="llm",
+        estimated=bool(structured.get("estimated", True)),
+    )
+
+
+@router.post("/fitness/water")
+def create_water(
+    payload: WaterCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return FitnessService(db, user.id).log_water(payload.amount_ml)
+
+
+@router.post("/fitness/weight")
+def create_weight(
+    payload: WeightCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    day = date.fromisoformat(payload.day) if payload.day else None
+    return FitnessService(db, user.id).log_weight(
+        payload.weight_kg,
+        body_fat_pct=payload.body_fat_pct,
+        chest_cm=payload.chest_cm,
+        waist_cm=payload.waist_cm,
+        neck_cm=payload.neck_cm,
+        hip_cm=payload.hip_cm,
+        notes=payload.notes,
+        recorded_at=payload.recorded_at,
+        day=day,
+        days_ago=payload.days_ago,
+    )
+
+
+@router.post("/fitness/body-composition/calc")
+def calc_body_composition(
+    payload: BodyCompositionCalc,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return FitnessService(db, user.id).calc_body_composition(payload.model_dump(exclude_none=True))
+
+
+@router.post("/fitness/steps")
+def create_steps(
+    payload: StepsCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    day = date.fromisoformat(payload.day) if payload.day else None
+    return FitnessService(db, user.id).log_steps(
+        payload.steps,
+        active_calories=payload.active_calories,
+        notes=payload.notes,
+        day=day,
+        days_ago=payload.days_ago,
+        logged_at=payload.logged_at,
+    )
+
+
+@router.delete("/fitness/steps/{log_id}")
+def delete_steps(log_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, bool]:
+    if not FitnessService(db, user.id).delete_step_log(log_id):
+        raise HTTPException(status_code=404, detail="Not found")
+    return {"ok": True}
+
+
+@router.post("/fitness/workouts")
+async def create_workout(
+    payload: WorkoutCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    service = FitnessService(db, user.id)
+    try:
+        structured = await structure_workout(payload.text)
+    except Exception as exc:
+        raise HTTPException(status_code=502, detail=str(exc)) from exc
+    day = date.fromisoformat(payload.day) if payload.day else None
+    return service.log_workout(
+        title=structured.get("title") or "Тренировка",
+        notes=structured.get("notes") or payload.text,
+        duration_min=structured.get("duration_min"),
+        exercises=structured.get("exercises"),
+        active_calories=structured.get("active_calories"),
+        total_calories=structured.get("total_calories"),
+        steps=structured.get("steps"),
+        day=day,
+        days_ago=payload.days_ago,
+        logged_at=payload.logged_at,
+    )
+
+
+@router.get("/fitness/body-metrics")
+def list_metrics(
+    limit: int = 30,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> list[dict[str, Any]]:
+    return FitnessService(db, user.id).list_body_metrics(limit=limit)
+
+
+@router.delete("/fitness/meals/{log_id}")
+def delete_meal(log_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, bool]:
+    if not FitnessService(db, user.id).delete_food_log(log_id):
+        raise HTTPException(status_code=404, detail="Not found")
+    return {"ok": True}
+
+
+@router.delete("/fitness/water/{log_id}")
+def delete_water(log_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, bool]:
+    if not FitnessService(db, user.id).delete_water_log(log_id):
+        raise HTTPException(status_code=404, detail="Not found")
+    return {"ok": True}
+
+
+@router.delete("/fitness/workouts/{log_id}")
+def delete_workout(log_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, bool]:
+    if not FitnessService(db, user.id).delete_workout_log(log_id):
+        raise HTTPException(status_code=404, detail="Not found")
+    return {"ok": True}
+
+
+@router.get("/fitness/reminders")
+def list_reminders(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[dict[str, Any]]:
+    return FitnessService(db, user.id).list_reminders()
+
+
+@router.put("/fitness/reminders/{kind}")
+def update_reminder(
+    kind: str,
+    payload: ReminderUpdate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return FitnessService(db, user.id).set_reminder(
+        kind,
+        enabled=payload.enabled,
+        hour=payload.hour,
+        minute=payload.minute,
+        interval_hours=payload.interval_hours,
+    )
+
+
+@router.get("/fitness/lookup/food")
+def lookup_food(q: str, limit: int = 5) -> list[dict[str, Any]]:
+    return OpenFoodFactsClient().search(q, limit=limit)
+
+
+@router.get("/fitness/lookup/exercise")
+def lookup_exercise(q: str, limit: int = 8) -> list[dict[str, Any]]:
+    return WgerClient().search_exercises(q, limit=limit)

backend/app/api/routes/memory.py

@@ -1,127 +1,130 @@
-from typing import Any
-
-from fastapi import APIRouter, Depends, HTTPException
-from pydantic import BaseModel, Field
-from sqlalchemy.orm import Session
-
-from app.db.base import get_db
-from app.db.models import ChatSession
-from app.memory.extract import extract_after_turn
-from app.memory.service import MemoryService
-
-router = APIRouter()
-
-
-class ProfileUpdate(BaseModel):
-    updates: dict[str, Any] = Field(default_factory=dict)
-
-
-class FactCreate(BaseModel):
-    content: str = Field(min_length=1)
-    category: str = "fact"
-    importance: int = Field(default=3, ge=1, le=5)
-    session_id: int | None = None
-
-
-class SessionSummaryUpdate(BaseModel):
-    summary: str = Field(min_length=1)
-    message_count: int = 0
-
-
-class ExtractRequest(BaseModel):
-    session_id: int
-    user_text: str = Field(min_length=1)
-    assistant_text: str = ""
-    force: bool = False
-
-
-@router.get("/memory")
-def get_memory_snapshot(
-    session_id: int | None = None,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    return MemoryService(db).snapshot(session_id)
-
-
-@router.get("/profile")
-def get_profile(db: Session = Depends(get_db)) -> dict[str, Any]:
-    return MemoryService(db).get_profile()
-
-
-@router.put("/profile")
-def update_profile(
-    payload: ProfileUpdate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return MemoryService(db).update_profile(payload.updates)
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.get("/memory/facts")
-def list_facts(
-    query: str | None = None,
-    category: str | None = None,
-    limit: int = 30,
-    db: Session = Depends(get_db),
-) -> list[dict[str, Any]]:
-    return MemoryService(db).recall_memories(query=query, category=category, limit=limit)
-
-
-@router.post("/memory/facts")
-def create_fact(
-    payload: FactCreate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return MemoryService(db).remember_fact(
-            payload.content,
-            category=payload.category,
-            session_id=payload.session_id,
-            importance=payload.importance,
-            source="api",
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.delete("/memory/facts/{memory_id}")
-def forget_fact(memory_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return MemoryService(db).forget_memory(memory_id)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
-
-
-@router.post("/memory/extract")
-async def extract_memories(
-    payload: ExtractRequest,
-    db: Session = Depends(get_db),
-) -> dict:
-    session = db.get(ChatSession, payload.session_id)
-    if not session:
-        raise HTTPException(status_code=404, detail="Session not found")
-    return await extract_after_turn(
-        db,
-        payload.session_id,
-        payload.user_text,
-        payload.assistant_text,
-        force=payload.force,
-    )
-
-
-@router.put("/memory/sessions/{session_id}/summary")
-def update_session_summary(
-    session_id: int,
-    payload: SessionSummaryUpdate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return MemoryService(db).update_session_summary(
-            session_id,
-            payload.summary,
-            message_count=payload.message_count,
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.db.models import ChatSession
+from app.memory.extract import extract_after_turn
+from app.memory.service import MemoryService
+
+router = APIRouter()
+
+
+class ProfileUpdate(BaseModel):
+    updates: dict[str, Any] = Field(default_factory=dict)
+
+
+class FactCreate(BaseModel):
+    content: str = Field(min_length=1)
+    category: str = "fact"
+    importance: int = Field(default=3, ge=1, le=5)
+    session_id: int | None = None
+
+
+class SessionSummaryUpdate(BaseModel):
+    summary: str = Field(min_length=1)
+    message_count: int = 0
+
+
+class ExtractRequest(BaseModel):
+    session_id: int
+    user_text: str = Field(min_length=1)
+    assistant_text: str = ""
+    force: bool = False
+
+
+@router.get("/memory")
+def get_memory_snapshot(
+    session_id: int | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    return MemoryService(db, user.id).snapshot(session_id)
+
+
+@router.get("/profile")
+def get_profile(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    return MemoryService(db, user.id).get_profile()
+
+
+@router.put("/profile")
+def update_profile(
+    payload: ProfileUpdate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return MemoryService(db, user.id).update_profile(payload.updates)
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.get("/memory/facts")
+def list_facts(
+    query: str | None = None,
+    category: str | None = None,
+    limit: int = 30,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> list[dict[str, Any]]:
+    return MemoryService(db, user.id).recall_memories(query=query, category=category, limit=limit)
+
+
+@router.post("/memory/facts")
+def create_fact(
+    payload: FactCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return MemoryService(db, user.id).remember_fact(
+            payload.content,
+            category=payload.category,
+            session_id=payload.session_id,
+            importance=payload.importance,
+            source="api",
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.delete("/memory/facts/{memory_id}")
+def forget_fact(memory_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return MemoryService(db, user.id).forget_memory(memory_id)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+
+
+@router.post("/memory/extract")
+async def extract_memories(
+    payload: ExtractRequest,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict:
+    session = db.get(ChatSession, payload.session_id)
+    if not session or session.user_id != user.id:
+        raise HTTPException(status_code=404, detail="Session not found")
+    return await extract_after_turn(
+        db,
+        payload.session_id,
+        payload.user_text,
+        payload.assistant_text,
+        user_id=user.id,
+        force=payload.force,
+    )
+
+
+@router.put("/memory/sessions/{session_id}/summary")
+def update_session_summary(
+    session_id: int,
+    payload: SessionSummaryUpdate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return MemoryService(db, user.id).update_session_summary(
+            session_id,
+            payload.summary,
+            message_count=payload.message_count,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc

backend/app/api/routes/pomodoro.py

@@ -1,100 +1,102 @@
-from fastapi import APIRouter, Depends, HTTPException
-from sqlalchemy.orm import Session
-
-from app.api.schemas import PomodoroStart, PomodoroStop
-from app.db.base import get_db
-from app.pomodoro.service import PomodoroService
-
-router = APIRouter()
-
-
-def _handle_value_error(exc: ValueError) -> HTTPException:
-    return HTTPException(status_code=400, detail=str(exc))
-
-
-@router.get("/status")
-def get_status(db: Session = Depends(get_db)) -> dict:
-    return PomodoroService(db).get_status()
-
-
-@router.post("/start")
-def start_pomodoro(payload: PomodoroStart, db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).start(
-            duration_min=payload.duration_min,
-            task_note=payload.task_note,
-        )
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.post("/pause")
-def pause_pomodoro(db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).pause()
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.post("/resume")
-def resume_pomodoro(db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).resume()
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.post("/stop")
-def stop_pomodoro(payload: PomodoroStop, db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).stop(
-            result=payload.result,
-            completed=payload.completed,
-        )
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.get("/history")
-def get_history(limit: int = 20, db: Session = Depends(get_db)) -> list[dict]:
-    return PomodoroService(db).history(limit=limit)
-
-
-@router.post("/work/start")
-def start_work(payload: PomodoroStart, db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).start_work(
-            duration_min=payload.duration_min,
-            task_note=payload.task_note,
-        )
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.post("/break/short/start")
-def start_short_break(duration_min: int | None = None, db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).start_short_break(duration_min=duration_min)
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.post("/break/long/start")
-def start_long_break(duration_min: int | None = None, db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).start_long_break(duration_min=duration_min)
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
-
-
-@router.post("/cycle/reset")
-def reset_cycle(clear_task: bool = False, db: Session = Depends(get_db)) -> dict:
-    return PomodoroService(db).reset_cycle(clear_task=clear_task)
-
-
-@router.post("/skip")
-def skip_phase(db: Session = Depends(get_db)) -> dict:
-    try:
-        return PomodoroService(db).skip_phase()
-    except ValueError as exc:
-        raise _handle_value_error(exc) from exc
+from fastapi import APIRouter, Depends, HTTPException
+from sqlalchemy.orm import Session
+
+from app.api.schemas import PomodoroStart, PomodoroStop
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.pomodoro.service import PomodoroService
+
+router = APIRouter()
+
+
+def _handle_value_error(exc: ValueError) -> HTTPException:
+    return HTTPException(status_code=400, detail=str(exc))
+
+
+@router.get("/status")
+def get_status(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    return PomodoroService(db, user.id).get_status()
+
+
+@router.post("/start")
+def start_pomodoro(payload: PomodoroStart, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).start(
+            duration_min=payload.duration_min,
+            task_note=payload.task_note,
+        )
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.post("/pause")
+def pause_pomodoro(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).pause()
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.post("/resume")
+def resume_pomodoro(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).resume()
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.post("/stop")
+def stop_pomodoro(payload: PomodoroStop, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).stop(
+            result=payload.result,
+            completed=payload.completed,
+        )
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.get("/history")
+def get_history(limit: int = 20, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[dict]:
+    return PomodoroService(db, user.id).history(limit=limit)
+
+
+@router.post("/work/start")
+def start_work(payload: PomodoroStart, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).start_work(
+            duration_min=payload.duration_min,
+            task_note=payload.task_note,
+        )
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.post("/break/short/start")
+def start_short_break(duration_min: int | None = None, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).start_short_break(duration_min=duration_min)
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.post("/break/long/start")
+def start_long_break(duration_min: int | None = None, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).start_long_break(duration_min=duration_min)
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc
+
+
+@router.post("/cycle/reset")
+def reset_cycle(clear_task: bool = False, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    return PomodoroService(db, user.id).reset_cycle(clear_task=clear_task)
+
+
+@router.post("/skip")
+def skip_phase(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict:
+    try:
+        return PomodoroService(db, user.id).skip_phase()
+    except ValueError as exc:
+        raise _handle_value_error(exc) from exc

backend/app/api/routes/projects.py

@@ -1,76 +1,78 @@
-from typing import Any
-
-from fastapi import APIRouter, Depends, HTTPException
-from pydantic import BaseModel, Field
-from sqlalchemy.orm import Session
-
-from app.db.base import get_db
-from app.projects.service import ProjectService
-
-router = APIRouter()
-
-
-class GiteaBinding(BaseModel):
-    gitea_owner: str = Field(min_length=1)
-    gitea_repo: str = Field(min_length=1)
-    default_branch: str = "main"
-
-
-class WorkItemCreate(BaseModel):
-    text: str = Field(min_length=1)
-    project_slug: str | None = None
-
-
-@router.get("/projects")
-def list_projects(db: Session = Depends(get_db)) -> list[dict[str, Any]]:
-    return ProjectService(db).list_projects()
-
-
-@router.post("/projects/sync-taiga")
-def sync_taiga_projects(db: Session = Depends(get_db)) -> list[dict[str, Any]]:
-    try:
-        return ProjectService(db).sync_taiga_projects()
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.put("/projects/{taiga_slug}/gitea")
-def bind_gitea(
-    taiga_slug: str,
-    payload: GiteaBinding,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return ProjectService(db).bind_gitea(
-            taiga_slug,
-            payload.gitea_owner,
-            payload.gitea_repo,
-            payload.default_branch,
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.post("/work-items")
-async def create_work_item(
-    payload: WorkItemCreate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return await ProjectService(db).create_work_item(
-            payload.text,
-            project_slug=payload.project_slug,
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-    except Exception as exc:
-        raise HTTPException(status_code=502, detail=str(exc)) from exc
-
-
-@router.get("/work-items")
-def list_work_items(
-    limit: int = 30,
-    status: str | None = None,
-    db: Session = Depends(get_db),
-) -> list[dict[str, Any]]:
-    return ProjectService(db).list_work_items(limit=limit, status=status)
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.projects.service import ProjectService
+
+router = APIRouter()
+
+
+class GiteaBinding(BaseModel):
+    gitea_owner: str = Field(min_length=1)
+    gitea_repo: str = Field(min_length=1)
+    default_branch: str = "main"
+
+
+class WorkItemCreate(BaseModel):
+    text: str = Field(min_length=1)
+    project_slug: str | None = None
+
+
+@router.get("/projects")
+def list_projects(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[dict[str, Any]]:
+    return ProjectService(db, user.id).list_projects()
+
+
+@router.post("/projects/sync-taiga")
+def sync_taiga_projects(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[dict[str, Any]]:
+    try:
+        return ProjectService(db, user.id).sync_taiga_projects()
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.put("/projects/{taiga_slug}/gitea")
+def bind_gitea(
+    taiga_slug: str,
+    payload: GiteaBinding,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return ProjectService(db, user.id).bind_gitea(
+            taiga_slug,
+            payload.gitea_owner,
+            payload.gitea_repo,
+            payload.default_branch,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.post("/work-items")
+async def create_work_item(
+    payload: WorkItemCreate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return await ProjectService(db, user.id).create_work_item(
+            payload.text,
+            project_slug=payload.project_slug,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+    except Exception as exc:
+        raise HTTPException(status_code=502, detail=str(exc)) from exc
+
+
+@router.get("/work-items")
+def list_work_items(
+    limit: int = 30,
+    status: str | None = None,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> list[dict[str, Any]]:
+    return ProjectService(db, user.id).list_work_items(limit=limit, status=status)

backend/app/api/routes/reminders.py

@@ -1,124 +1,128 @@
-from datetime import datetime, timezone
-from typing import Any
-from zoneinfo import ZoneInfo
-
-from fastapi import APIRouter, Depends, HTTPException, Query
-from pydantic import BaseModel, Field
-from sqlalchemy.orm import Session
-
-from app.db.base import get_db
-from app.homelab.context import resolve_timezone
-from app.reminders.service import RemindersService
-
-router = APIRouter()
-
-
-class ReminderCreate(BaseModel):
-    title: str = Field(min_length=1, max_length=255)
-    due_at: str = Field(description="ISO datetime, например 2027-05-12T12:16:00")
-    notes: str = ""
-    all_day: bool = False
-    recurrence: str = "none"
-
-
-class ReminderUpdate(BaseModel):
-    title: str | None = Field(default=None, min_length=1, max_length=255)
-    due_at: str | None = None
-    notes: str | None = None
-    all_day: bool | None = None
-    recurrence: str | None = None
-    enabled: bool | None = None
-
-
-@router.get("")
-def get_snapshot(db: Session = Depends(get_db)) -> dict[str, Any]:
-    return RemindersService(db).snapshot()
-
-
-@router.get("/upcoming")
-def list_upcoming(
-    limit: int = Query(30, ge=1, le=100),
-    db: Session = Depends(get_db),
-) -> list[dict[str, Any]]:
-    return RemindersService(db).list_upcoming(limit=limit)
-
-
-@router.get("/calendar")
-def calendar(
-    year: int = Query(..., ge=2000, le=2100),
-    month: int = Query(..., ge=1, le=12),
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    tz_name = resolve_timezone(db)
-    try:
-        tz = ZoneInfo(tz_name)
-    except Exception:
-        tz = ZoneInfo("Europe/Moscow")
-
-    start = datetime(year, month, 1, tzinfo=tz)
-    if month == 12:
-        end = datetime(year + 1, 1, 1, tzinfo=tz)
-    else:
-        end = datetime(year, month + 1, 1, tzinfo=tz)
-
-    service = RemindersService(db)
-    items = service.list_in_range(
-        date_from=start.astimezone(timezone.utc),
-        date_to=end.astimezone(timezone.utc),
-    )
-    return {
-        "year": year,
-        "month": month,
-        "timezone": tz_name,
-        "reminders": items,
-    }
-
-
-@router.post("")
-def create_reminder(payload: ReminderCreate, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return RemindersService(db).create(
-            title=payload.title,
-            due_at=payload.due_at,
-            notes=payload.notes,
-            all_day=payload.all_day,
-            recurrence=payload.recurrence,
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.patch("/{reminder_id}")
-def update_reminder(
-    reminder_id: int,
-    payload: ReminderUpdate,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return RemindersService(db).update(
-            reminder_id,
-            title=payload.title,
-            due_at=payload.due_at,
-            notes=payload.notes,
-            all_day=payload.all_day,
-            recurrence=payload.recurrence,
-            enabled=payload.enabled,
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
-
-
-@router.delete("/{reminder_id}")
-def delete_reminder(reminder_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return RemindersService(db).delete(reminder_id)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
-
-
-@router.post("/{reminder_id}/complete")
-def complete_reminder(reminder_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return RemindersService(db).complete(reminder_id)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
+from datetime import datetime, timezone
+from typing import Any
+from zoneinfo import ZoneInfo
+
+from fastapi import APIRouter, Depends, HTTPException, Query
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.homelab.context import resolve_timezone
+from app.reminders_scoped.service import RemindersService
+
+router = APIRouter()
+
+
+class ReminderCreate(BaseModel):
+    title: str = Field(min_length=1, max_length=255)
+    due_at: str = Field(description="ISO datetime, например 2027-05-12T12:16:00")
+    notes: str = ""
+    all_day: bool = False
+    recurrence: str = "none"
+
+
+class ReminderUpdate(BaseModel):
+    title: str | None = Field(default=None, min_length=1, max_length=255)
+    due_at: str | None = None
+    notes: str | None = None
+    all_day: bool | None = None
+    recurrence: str | None = None
+    enabled: bool | None = None
+
+
+@router.get("")
+def get_snapshot(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    return RemindersService(db, user.id).snapshot()
+
+
+@router.get("/upcoming")
+def list_upcoming(
+    limit: int = Query(30, ge=1, le=100),
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> list[dict[str, Any]]:
+    return RemindersService(db, user.id).list_upcoming(limit=limit)
+
+
+@router.get("/calendar")
+def calendar(
+    year: int = Query(..., ge=2000, le=2100),
+    month: int = Query(..., ge=1, le=12),
+    db: Session = Depends(get_db),
+    user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    tz_name = resolve_timezone(db, user.id)
+    try:
+        tz = ZoneInfo(tz_name)
+    except Exception:
+        tz = ZoneInfo("Europe/Moscow")
+
+    start = datetime(year, month, 1, tzinfo=tz)
+    if month == 12:
+        end = datetime(year + 1, 1, 1, tzinfo=tz)
+    else:
+        end = datetime(year, month + 1, 1, tzinfo=tz)
+
+    service = RemindersService(db, user.id)
+    items = service.list_in_range(
+        date_from=start.astimezone(timezone.utc),
+        date_to=end.astimezone(timezone.utc),
+    )
+    return {
+        "year": year,
+        "month": month,
+        "timezone": tz_name,
+        "reminders": items,
+    }
+
+
+@router.post("")
+def create_reminder(payload: ReminderCreate, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return RemindersService(db, user.id).create(
+            title=payload.title,
+            due_at=payload.due_at,
+            notes=payload.notes,
+            all_day=payload.all_day,
+            recurrence=payload.recurrence,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.patch("/{reminder_id}")
+def update_reminder(
+    reminder_id: int,
+    payload: ReminderUpdate,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return RemindersService(db, user.id).update(
+            reminder_id,
+            title=payload.title,
+            due_at=payload.due_at,
+            notes=payload.notes,
+            all_day=payload.all_day,
+            recurrence=payload.recurrence,
+            enabled=payload.enabled,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+
+
+@router.delete("/{reminder_id}")
+def delete_reminder(reminder_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return RemindersService(db, user.id).delete(reminder_id)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+
+
+@router.post("/{reminder_id}/complete")
+def complete_reminder(reminder_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return RemindersService(db, user.id).complete(reminder_id)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc

backend/app/api/routes/settings.py

@@ -0,0 +1,34 @@
+from typing import Any
+
+from fastapi import APIRouter, Depends
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.settings.service import SETTING_KEYS, SettingsService
+
+router = APIRouter()
+
+
+class SettingsPatch(BaseModel):
+    openrouter_model: str | None = None
+    memory_extract_model: str | None = None
+    openrouter_reasoning_effort: str | None = None
+    rag_enabled: bool | None = None
+    rag_top_k: int | None = Field(default=None, ge=1, le=50)
+
+
+@router.get("/settings")
+def get_settings_route(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    return SettingsService(db).snapshot()
+
+
+@router.patch("/settings")
+def patch_settings_route(
+    payload: SettingsPatch,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    updates = payload.model_dump(exclude_unset=True)
+    return SettingsService(db).patch(updates)

backend/app/api/routes/shopping.py

@@ -1,116 +1,118 @@
-from typing import Any
-
-from fastapi import APIRouter, Depends, HTTPException
-from pydantic import BaseModel, Field
-from sqlalchemy.orm import Session
-
-from app.db.base import get_db
-from app.shopping.service import ShoppingService
-
-router = APIRouter()
-
-
-class ListCreate(BaseModel):
-    name: str = Field(min_length=1, max_length=255)
-
-
-class ListRename(BaseModel):
-    name: str = Field(min_length=1, max_length=255)
-
-
-class ItemInput(BaseModel):
-    text: str = Field(min_length=1, max_length=500)
-    quantity: float | None = None
-    unit: str = ""
-
-
-class ItemsAdd(BaseModel):
-    list_id: int | None = None
-    list_name: str | None = None
-    items: list[ItemInput] = Field(min_length=1)
-
-
-class ItemChecked(BaseModel):
-    checked: bool
-
-
-@router.get("")
-def get_snapshot(db: Session = Depends(get_db)) -> dict[str, Any]:
-    return ShoppingService(db).snapshot()
-
-
-@router.get("/lists")
-def list_lists(db: Session = Depends(get_db)) -> list[dict[str, Any]]:
-    return ShoppingService(db).list_lists(include_items=True)
-
-
-@router.post("/lists")
-def create_list(payload: ListCreate, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).create_list(payload.name)
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.get("/lists/{list_id}")
-def get_list(list_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    data = ShoppingService(db).get_list(list_id=list_id)
-    if not data:
-        raise HTTPException(status_code=404, detail="List not found")
-    return data
-
-
-@router.patch("/lists/{list_id}")
-def rename_list(list_id: int, payload: ListRename, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).rename_list(list_id, payload.name)
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.delete("/lists/{list_id}")
-def delete_list(list_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).delete_list(list_id)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
-
-
-@router.post("/items")
-def add_items(payload: ItemsAdd, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).add_items(
-            [i.model_dump() for i in payload.items],
-            list_id=payload.list_id,
-            list_name=payload.list_name,
-        )
-    except ValueError as exc:
-        raise HTTPException(status_code=400, detail=str(exc)) from exc
-
-
-@router.patch("/items/{item_id}")
-def set_item_checked(
-    item_id: int,
-    payload: ItemChecked,
-    db: Session = Depends(get_db),
-) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).set_item_checked(item_id, payload.checked)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
-
-
-@router.delete("/items/{item_id}")
-def remove_item(item_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).remove_item(item_id)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
-
-
-@router.post("/lists/{list_id}/clear-checked")
-def clear_checked(list_id: int, db: Session = Depends(get_db)) -> dict[str, Any]:
-    try:
-        return ShoppingService(db).clear_checked(list_id)
-    except ValueError as exc:
-        raise HTTPException(status_code=404, detail=str(exc)) from exc
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException
+from pydantic import BaseModel, Field
+from sqlalchemy.orm import Session
+
+from app.auth.deps import get_current_user
+from app.db.base import get_db
+from app.db.models import User
+from app.shopping.service import ShoppingService
+
+router = APIRouter()
+
+
+class ListCreate(BaseModel):
+    name: str = Field(min_length=1, max_length=255)
+
+
+class ListRename(BaseModel):
+    name: str = Field(min_length=1, max_length=255)
+
+
+class ItemInput(BaseModel):
+    text: str = Field(min_length=1, max_length=500)
+    quantity: float | None = None
+    unit: str = ""
+
+
+class ItemsAdd(BaseModel):
+    list_id: int | None = None
+    list_name: str | None = None
+    items: list[ItemInput] = Field(min_length=1)
+
+
+class ItemChecked(BaseModel):
+    checked: bool
+
+
+@router.get("")
+def get_snapshot(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    return ShoppingService(db, user.id).snapshot()
+
+
+@router.get("/lists")
+def list_lists(db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> list[dict[str, Any]]:
+    return ShoppingService(db, user.id).list_lists(include_items=True)
+
+
+@router.post("/lists")
+def create_list(payload: ListCreate, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).create_list(payload.name)
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.get("/lists/{list_id}")
+def get_list(list_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    data = ShoppingService(db, user.id).get_list(list_id=list_id)
+    if not data:
+        raise HTTPException(status_code=404, detail="List not found")
+    return data
+
+
+@router.patch("/lists/{list_id}")
+def rename_list(list_id: int, payload: ListRename, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).rename_list(list_id, payload.name)
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.delete("/lists/{list_id}")
+def delete_list(list_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).delete_list(list_id)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+
+
+@router.post("/items")
+def add_items(payload: ItemsAdd, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).add_items(
+            [i.model_dump() for i in payload.items],
+            list_id=payload.list_id,
+            list_name=payload.list_name,
+        )
+    except ValueError as exc:
+        raise HTTPException(status_code=400, detail=str(exc)) from exc
+
+
+@router.patch("/items/{item_id}")
+def set_item_checked(
+    item_id: int,
+    payload: ItemChecked,
+    db: Session = Depends(get_db), user: User = Depends(get_current_user),
+) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).set_item_checked(item_id, payload.checked)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+
+
+@router.delete("/items/{item_id}")
+def remove_item(item_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).remove_item(item_id)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc
+
+
+@router.post("/lists/{list_id}/clear-checked")
+def clear_checked(list_id: int, db: Session = Depends(get_db), user: User = Depends(get_current_user)) -> dict[str, Any]:
+    try:
+        return ShoppingService(db, user.id).clear_checked(list_id)
+    except ValueError as exc:
+        raise HTTPException(status_code=404, detail=str(exc)) from exc

backend/app/api/routes/webhooks.py

@@ -1,118 +1,106 @@
-import hashlib
-import hmac
-import json
-import logging
-from typing import Any
-
-from fastapi import APIRouter, Depends, HTTPException, Request
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.config import get_settings
-from app.db.base import SessionLocal, get_db
-from app.db.models import ChatSession, Message, ProjectBinding
-from app.projects.service import ProjectService
-
-router = APIRouter()
-logger = logging.getLogger(__name__)
-
-
-def _verify_gitea_signature(body: bytes, signature: str | None, secret: str) -> bool:
-    if not secret:
-        return True
-    if not signature:
-        return False
-    if signature.startswith("sha256="):
-        signature = signature[7:]
-    expected = hmac.new(secret.encode(), body, hashlib.sha256).hexdigest()
-    return hmac.compare_digest(expected, signature)
-
-
-def _post_close_notice(results: list[dict[str, Any]], owner: str, repo: str) -> None:
-    if not results:
-        return
-    db = SessionLocal()
-    try:
-        session = db.scalar(
-            select(ChatSession).order_by(ChatSession.updated_at.desc()).limit(1)
-        )
-        if not session:
-            session = ChatSession(title="Git")
-            db.add(session)
-            db.commit()
-            db.refresh(session)
-
-        lines = [f"🔀 **Push** `{owner}/{repo}`"]
-        for item in results:
-            if "closed" in item:
-                lines.append(f"- `{item.get('commit', '?')}`: закрыто {item['closed']}")
-            elif "error" in item:
-                lines.append(f"- ошибка: {item['error']}")
-
-        db.add(Message(session_id=session.id, role="notice", content="\n".join(lines)))
-        db.commit()
-    finally:
-        db.close()
-
-
-@router.post("/webhooks/gitea")
-async def gitea_webhook(request: Request, db: Session = Depends(get_db)) -> dict[str, Any]:
-    body = await request.body()
-    settings = get_settings()
-    signature = (
-        request.headers.get("X-Gitea-Signature")
-        or request.headers.get("X-Gogs-Signature")
-        or request.headers.get("X-Hub-Signature-256")
-    )
-
-    if not _verify_gitea_signature(body, signature, settings.gitea_webhook_secret):
-        raise HTTPException(status_code=401, detail="Invalid webhook signature")
-
-    payload = json.loads(body)
-    if payload.get("secret") and settings.gitea_webhook_secret:
-        if payload.get("secret") != settings.gitea_webhook_secret:
-            raise HTTPException(status_code=401, detail="Invalid webhook secret")
-
-    event = request.headers.get("X-Gitea-Event", "")
-    if event != "push":
-        return {"ok": True, "skipped": event}
-
-    repo = payload.get("repository", {})
-    owner = repo.get("owner", {}).get("login", "")
-    repo_name = repo.get("name", "")
-    if not owner or not repo_name:
-        raise HTTPException(status_code=400, detail="Missing repository info")
-
-    binding = db.scalar(
-        select(ProjectBinding).where(
-            ProjectBinding.gitea_owner == owner,
-            ProjectBinding.gitea_repo == repo_name,
-        )
-    )
-    if not binding:
-        return {"ok": True, "skipped": "unknown repo"}
-
-    commits = list(payload.get("commits") or [])
-    if not commits:
-        head = payload.get("head_commit")
-        if head:
-            commits = [head]
-
-    logger.info(
-        "Gitea push %s/%s ref=%s commits=%d",
-        owner,
-        repo_name,
-        payload.get("ref", ""),
-        len(commits),
-    )
-
-    service = ProjectService(db)
-    results = service.process_push(owner, repo_name, commits)
-    if results:
-        logger.info("Gitea push results: %s", results)
-    else:
-        logger.warning("Gitea push: no close actions for %s/%s", owner, repo_name)
-
-    _post_close_notice(results, owner, repo_name)
-
-    return {"ok": True, "results": results, "commits_processed": len(commits)}
+import hashlib
+import hmac
+import json
+import logging
+from typing import Any
+
+from fastapi import APIRouter, Depends, HTTPException, Request
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.chat.notice_inbox import post_notice_to_latest_chat
+from app.config import get_settings
+from app.db.base import get_db
+from app.db.models import ProjectBinding
+from app.projects.service import ProjectService
+
+router = APIRouter()
+logger = logging.getLogger(__name__)
+
+
+def _verify_gitea_signature(body: bytes, signature: str | None, secret: str) -> bool:
+    if not secret:
+        return True
+    if not signature:
+        return False
+    if signature.startswith("sha256="):
+        signature = signature[7:]
+    expected = hmac.new(secret.encode(), body, hashlib.sha256).hexdigest()
+    return hmac.compare_digest(expected, signature)
+
+
+def _post_close_notice(
+    results: list[dict[str, Any]], owner: str, repo: str, user_id: int
+) -> None:
+    if not results:
+        return
+    lines = [f"🔀 **Push** `{owner}/{repo}`"]
+    for item in results:
+        if "closed" in item:
+            lines.append(f"- `{item.get('commit', '?')}`: закрыто {item['closed']}")
+        elif "error" in item:
+            lines.append(f"- ошибка: {item['error']}")
+    post_notice_to_latest_chat("\n".join(lines), user_id)
+
+
+@router.post("/webhooks/gitea")
+async def gitea_webhook(request: Request, db: Session = Depends(get_db)) -> dict[str, Any]:
+    body = await request.body()
+    settings = get_settings()
+    signature = (
+        request.headers.get("X-Gitea-Signature")
+        or request.headers.get("X-Gogs-Signature")
+        or request.headers.get("X-Hub-Signature-256")
+    )
+
+    if not _verify_gitea_signature(body, signature, settings.gitea_webhook_secret):
+        raise HTTPException(status_code=401, detail="Invalid webhook signature")
+
+    payload = json.loads(body)
+    if payload.get("secret") and settings.gitea_webhook_secret:
+        if payload.get("secret") != settings.gitea_webhook_secret:
+            raise HTTPException(status_code=401, detail="Invalid webhook secret")
+
+    event = request.headers.get("X-Gitea-Event", "")
+    if event != "push":
+        return {"ok": True, "skipped": event}
+
+    repo = payload.get("repository", {})
+    owner = repo.get("owner", {}).get("login", "")
+    repo_name = repo.get("name", "")
+    if not owner or not repo_name:
+        raise HTTPException(status_code=400, detail="Missing repository info")
+
+    binding = db.scalar(
+        select(ProjectBinding).where(
+            ProjectBinding.gitea_owner == owner,
+            ProjectBinding.gitea_repo == repo_name,
+        )
+    )
+    if not binding:
+        return {"ok": True, "skipped": "unknown repo"}
+
+    commits = list(payload.get("commits") or [])
+    if not commits:
+        head = payload.get("head_commit")
+        if head:
+            commits = [head]
+
+    logger.info(
+        "Gitea push %s/%s ref=%s commits=%d",
+        owner,
+        repo_name,
+        payload.get("ref", ""),
+        len(commits),
+    )
+
+    service = ProjectService(db, binding.user_id)
+    results = service.process_push(owner, repo_name, commits)
+    if results:
+        logger.info("Gitea push results: %s", results)
+    else:
+        logger.warning("Gitea push: no close actions for %s/%s", owner, repo_name)
+
+    _post_close_notice(results, owner, repo_name, binding.user_id)
+
+    return {"ok": True, "results": results, "commits_processed": len(commits)}

backend/app/auth/__init__.py

@@ -0,0 +1,5 @@
+from app.auth.deps import get_current_user
+from app.auth.service import create_user, find_user_by_token
+from app.auth.tokens import hash_token, verify_token
+
+__all__ = ["get_current_user", "hash_token", "verify_token"]

backend/app/auth/deps.py

@@ -0,0 +1,34 @@
+from fastapi import Depends, HTTPException, Request, status
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.auth.tokens import hash_token
+from app.db.base import get_db
+from app.db.models import User
+
+
+def _extract_token(request: Request) -> str | None:
+    auth = request.headers.get("Authorization", "")
+    if auth.lower().startswith("bearer "):
+        token = auth[7:].strip()
+        if token:
+            return token
+    header = request.headers.get("X-API-Token", "").strip()
+    return header or None
+
+
+def get_current_user(
+    request: Request,
+    db: Session = Depends(get_db),
+) -> User:
+    token = _extract_token(request)
+    if not token:
+        raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="Missing API token")
+
+    token_hash = hash_token(token)
+    user = db.scalar(
+        select(User).where(User.api_token_hash == token_hash, User.is_active.is_(True))
+    )
+    if not user:
+        raise HTTPException(status_code=status.HTTP_401_UNAUTHORIZED, detail="Invalid API token")
+    return user

backend/app/auth/service.py

@@ -0,0 +1,61 @@
+import secrets
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.auth.tokens import hash_token
+from app.db.models import CharacterCard, User
+from app.character.card import DEFAULT_CARD, normalize_card
+import json
+
+
+def find_user_by_token(db: Session, token: str) -> User | None:
+    token_hash = hash_token(token.strip())
+    return db.scalar(
+        select(User).where(User.api_token_hash == token_hash, User.is_active.is_(True))
+    )
+
+
+def user_to_dict(user: User) -> dict[str, Any]:
+    return {
+        "id": user.id,
+        "username": user.username,
+        "display_name": user.display_name or user.username,
+    }
+
+
+def create_user(
+    db: Session,
+    *,
+    username: str,
+    display_name: str = "",
+    api_token: str | None = None,
+) -> tuple[User, str]:
+    clean = username.strip().lower()
+    if not clean:
+        raise ValueError("username не может быть пустым")
+    existing = db.scalar(select(User).where(User.username == clean))
+    if existing:
+        raise ValueError(f"Пользователь «{clean}» уже существует")
+
+    plain_token = (api_token or "").strip() or secrets.token_urlsafe(32)
+    user = User(
+        username=clean,
+        display_name=(display_name or clean).strip(),
+        api_token_hash=hash_token(plain_token),
+        is_active=True,
+    )
+    db.add(user)
+    db.flush()
+
+    card = normalize_card(DEFAULT_CARD)
+    db.add(
+        CharacterCard(
+            user_id=user.id,
+            card_json=json.dumps(card, ensure_ascii=False),
+        )
+    )
+    db.commit()
+    db.refresh(user)
+    return user, plain_token

backend/app/auth/tokens.py

@@ -0,0 +1,9 @@
+import hashlib
+
+
+def hash_token(token: str) -> str:
+    return hashlib.sha256(token.encode("utf-8")).hexdigest()
+
+
+def verify_token(plain: str, token_hash: str) -> bool:
+    return hash_token(plain) == token_hash

backend/app/character/card.py

@@ -1,100 +1,100 @@
-from typing import Any
-
-TOOLS_INSTRUCTIONS = """
-Ты также домашний ассистент с инструментами помидоро-цикла (работа → перерыв → работа → длинный перерыв → сброс).
-Обязательные правила:
-- Любой вопрос о таймере, помидоро, задачах или истории — СНАЧАЛА вызывай соответствующий инструмент.
-- Никогда не выдумывай статус таймера или список задач.
-- После вызова инструмента кратко объясни результат пользователю по-человечески.
-- Помидоро: get_pomodoro_status, start_pomodoro, start_short_break, start_long_break,
-  stop_pomodoro, skip_pomodoro_phase, reset_pomodoro_cycle, get_pomodoro_history.
-- Taiga: sync_taiga_projects, list_taiga_projects, list_taiga_tasks, create_work_item, list_work_items.
-- «Какие задачи» / «покажи задачи проекта» → list_taiga_tasks (живые данные Taiga).
-- list_work_items — ТОЛЬКО задачи, созданные через create_work_item (локальная БД).
-- create_work_item — при «заведи баг/фичу»; передай полный текст и project_slug.
-- Фитнес: get_fitness_summary (date/days_ago), get_fitness_history, set_fitness_profile, log_meal, log_water, log_weight, log_workout,
-- «Что ел вчера» → get_fitness_summary days_ago=1. «За неделю» → get_fitness_history.
-  calc_fitness_targets, lookup_food, lookup_exercise, set_fitness_reminder.
-- Память: remember_fact, recall_memories, forget_memory, update_profile, update_session_summary.
-- «Запомни» → remember_fact. «Кто я» / «сколько мне лет» → профиль и факты из блока [Память], не выдумывай.
-- Сценарий персонажа (сын, семья) — тон общения, НЕ факты о пользователе.
-- Снимок проектов/задач и памяти есть в контексте, но для записи/поиска вызывай tools.
-- Никогда не пиши «ожидаю ответа от системы».
-- В текстовых ответах пользователю не используй эмодзи.
-- Погода: get_weather или блок [Погода] в контексте; «что на улице» / «будет ли дождь» — не выдумывай.
-- Утренний брифинг (погода + новости) → get_morning_briefing.
-- Картинки: generate_image — «нарисуй себя» → draw_self=true; иначе scene_description на английском (booru-теги). Внешность из карточки персонажа. Не злоупотребляй.
-- Покупки: list_shopping_lists, create_shopping_list, add_shopping_items, check_shopping_item, remove_shopping_item, delete_shopping_list.
-- «Добавь в список покупок» → add_shopping_items (list_name + товары). «Что купить» → list_shopping_lists. Не выдумывай списки.
-- Напоминания: list_reminders, create_reminder, update_reminder, delete_reminder, complete_reminder.
-- «Напомни через 15 минут», «завтра утром», «12 мая в 9:00» → create_reminder с due_at в ISO (часовой пояс из [Текущее время]).
-- День рождения, Новый год и другие праздники → recurrence yearly.
-- Относительное время считай от «Сейчас» в контексте. «Утром» ≈ 09:00, «вечером» ≈ 19:00, если не уточнено иначе.
-""".strip()
-
-DEFAULT_CARD: dict[str, Any] = {
-    "spec": "chara_card_v2",
-    "spec_version": "2.0",
-    "data": {
-        "name": "Домашний ассистент",
-        "description": "Дружелюбный ИИ-помощник для дома. Отвечает на вопросы, даёт советы, помогает с помидоро-таймером.",
-        "personality": "Тёплый, остроумный, по делу. Говорит на русском. Может шутить, но не перегибает.",
-        "scenario": "Пользователь общается с ассистентом дома через веб-интерфейс.",
-        "first_mes": "Привет! Чем займёмся — поболтаем или заведём помидоро?",
-        "mes_example": "",
-        "system_prompt": "",
-        "post_history_instructions": "",
-        "alternate_greetings": [],
-        "tags": ["assistant", "home", "pomodoro"],
-        "creator": "",
-        "creator_notes": "",
-        "character_version": "1.0",
-        "appearance_tags": "",
-        "appearance_prose": "",
-        "lora_name": "",
-        "lora_weight": 0.8,
-        "rp_persona_id": "",
-        "sd_enabled": True,
-    },
-}
-
-
-def normalize_card(raw: dict[str, Any]) -> dict[str, Any]:
-    if "data" in raw and isinstance(raw["data"], dict):
-        card = {
-            "spec": raw.get("spec", "chara_card_v2"),
-            "spec_version": raw.get("spec_version", "2.0"),
-            "data": {**DEFAULT_CARD["data"], **raw["data"]},
-        }
-        return card
-
-    if "name" in raw or "description" in raw:
-        return {
-            "spec": "chara_card_v2",
-            "spec_version": "2.0",
-            "data": {**DEFAULT_CARD["data"], **raw},
-        }
-
-    return DEFAULT_CARD.copy()
-
-
-def build_system_prompt(card: dict[str, Any]) -> str:
-    data = card.get("data", {})
-    parts: list[str] = []
-
-    name = data.get("name", "Ассистент")
-    parts.append(f"Ты — {name}.")
-
-    if data.get("system_prompt"):
-        parts.append(data["system_prompt"])
-    if data.get("description"):
-        parts.append(data["description"])
-    if data.get("personality"):
-        parts.append(f"Характер: {data['personality']}")
-    if data.get("scenario"):
-        parts.append(f"Сценарий: {data['scenario']}")
-    if data.get("post_history_instructions"):
-        parts.append(data["post_history_instructions"])
-
-    parts.append(TOOLS_INSTRUCTIONS)
-    return "\n\n".join(part for part in parts if part.strip())
+from typing import Any
+
+TOOLS_INSTRUCTIONS = """
+Ты также домашний ассистент с инструментами помидоро-цикла (работа → перерыв → работа → длинный перерыв → сброс).
+Обязательные правила:
+- Любой вопрос о таймере, помидоро, задачах или истории — СНАЧАЛА вызывай соответствующий инструмент.
+- Никогда не выдумывай статус таймера или список задач.
+- После вызова инструмента кратко объясни результат пользователю по-человечески.
+- Помидоро: get_pomodoro_status, start_pomodoro, start_short_break, start_long_break,
+  stop_pomodoro, skip_pomodoro_phase, reset_pomodoro_cycle, get_pomodoro_history.
+- Taiga: sync_taiga_projects, list_taiga_projects, list_taiga_tasks, create_work_item, list_work_items.
+- «Какие задачи» / «покажи задачи проекта» → list_taiga_tasks (живые данные Taiga).
+- list_work_items — ТОЛЬКО задачи, созданные через create_work_item (локальная БД).
+- create_work_item — при «заведи баг/фичу»; передай полный текст и project_slug.
+- Фитнес: get_fitness_summary (date/days_ago), get_fitness_history, set_fitness_profile, log_meal, log_water, log_weight (neck_cm/waist_cm/hip_cm → Navy), log_workout,
+- «Что ел вчера» → get_fitness_summary days_ago=1. «За неделю» → get_fitness_history.
+  calc_fitness_targets, calc_body_composition (расчёт Navy/WHR/LBM/FFMI без записи), lookup_food, lookup_exercise, set_fitness_reminder.
+- Память: remember_fact, recall_memories, forget_memory, update_profile, update_session_summary.
+- «Запомни» → remember_fact. «Кто я» / «сколько мне лет» → профиль и факты из блока [Память], не выдумывай.
+- Сценарий персонажа (сын, семья) — тон общения, НЕ факты о пользователе.
+- Снимок проектов/задач и памяти есть в контексте, но для записи/поиска вызывай tools.
+- Никогда не пиши «ожидаю ответа от системы».
+- В текстовых ответах пользователю не используй эмодзи.
+- Погода: get_weather или блок [Погода] в контексте; «что на улице» / «будет ли дождь» — не выдумывай.
+- Утренний брифинг (погода + новости) → get_morning_briefing.
+- Картинки: generate_image — «нарисуй себя» → draw_self=true; иначе scene_description на английском (booru-теги). Внешность из карточки персонажа. Не злоупотребляй.
+- Покупки: list_shopping_lists, create_shopping_list, add_shopping_items, check_shopping_item, remove_shopping_item, delete_shopping_list.
+- «Добавь в список покупок» → add_shopping_items (list_name + товары). «Что купить» → list_shopping_lists. Не выдумывай списки.
+- Напоминания: list_reminders, create_reminder, update_reminder, delete_reminder, complete_reminder.
+- «Напомни через 15 минут», «завтра утром», «12 мая в 9:00» → create_reminder с due_at в ISO (часовой пояс из [Текущее время]).
+- День рождения, Новый год и другие праздники → recurrence yearly.
+- Относительное время считай от «Сейчас» в контексте. «Утром» ≈ 09:00, «вечером» ≈ 19:00, если не уточнено иначе.
+""".strip()
+
+DEFAULT_CARD: dict[str, Any] = {
+    "spec": "chara_card_v2",
+    "spec_version": "2.0",
+    "data": {
+        "name": "Домашний ассистент",
+        "description": "Дружелюбный ИИ-помощник для дома. Отвечает на вопросы, даёт советы, помогает с помидоро-таймером.",
+        "personality": "Тёплый, остроумный, по делу. Говорит на русском. Может шутить, но не перегибает.",
+        "scenario": "Пользователь общается с ассистентом дома через веб-интерфейс.",
+        "first_mes": "Привет! Чем займёмся — поболтаем или заведём помидоро?",
+        "mes_example": "",
+        "system_prompt": "",
+        "post_history_instructions": "",
+        "alternate_greetings": [],
+        "tags": ["assistant", "home", "pomodoro"],
+        "creator": "",
+        "creator_notes": "",
+        "character_version": "1.0",
+        "appearance_tags": "",
+        "appearance_prose": "",
+        "lora_name": "",
+        "lora_weight": 0.8,
+        "rp_persona_id": "",
+        "sd_enabled": True,
+    },
+}
+
+
+def normalize_card(raw: dict[str, Any]) -> dict[str, Any]:
+    if "data" in raw and isinstance(raw["data"], dict):
+        card = {
+            "spec": raw.get("spec", "chara_card_v2"),
+            "spec_version": raw.get("spec_version", "2.0"),
+            "data": {**DEFAULT_CARD["data"], **raw["data"]},
+        }
+        return card
+
+    if "name" in raw or "description" in raw:
+        return {
+            "spec": "chara_card_v2",
+            "spec_version": "2.0",
+            "data": {**DEFAULT_CARD["data"], **raw},
+        }
+
+    return DEFAULT_CARD.copy()
+
+
+def build_system_prompt(card: dict[str, Any]) -> str:
+    data = card.get("data", {})
+    parts: list[str] = []
+
+    name = data.get("name", "Ассистент")
+    parts.append(f"Ты — {name}.")
+
+    if data.get("system_prompt"):
+        parts.append(data["system_prompt"])
+    if data.get("description"):
+        parts.append(data["description"])
+    if data.get("personality"):
+        parts.append(f"Характер: {data['personality']}")
+    if data.get("scenario"):
+        parts.append(f"Сценарий: {data['scenario']}")
+    if data.get("post_history_instructions"):
+        parts.append(data["post_history_instructions"])
+
+    parts.append(TOOLS_INSTRUCTIONS)
+    return "\n\n".join(part for part in parts if part.strip())

backend/app/character/service.py

@@ -1,27 +1,43 @@
-import json
-from pathlib import Path
-from typing import Any
-
-from app.character.card import DEFAULT_CARD, build_system_prompt, normalize_card
-
-CARD_PATH = Path("./data/character.json")
-
-
-class CharacterService:
-    def get_card(self) -> dict[str, Any]:
-        if CARD_PATH.is_file():
-            try:
-                raw = json.loads(CARD_PATH.read_text(encoding="utf-8"))
-                return normalize_card(raw)
-            except (json.JSONDecodeError, OSError):
-                pass
-        return normalize_card(DEFAULT_CARD)
-
-    def save_card(self, raw: dict[str, Any]) -> dict[str, Any]:
-        card = normalize_card(raw)
-        CARD_PATH.parent.mkdir(parents=True, exist_ok=True)
-        CARD_PATH.write_text(json.dumps(card, ensure_ascii=False, indent=2), encoding="utf-8")
-        return card
-
-    def get_system_prompt(self) -> str:
-        return build_system_prompt(self.get_card())
+import json
+from datetime import datetime, timezone
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.character.card import DEFAULT_CARD, build_system_prompt, normalize_card
+from app.db.models import CharacterCard
+
+
+class CharacterService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+
+    def get_card(self) -> dict[str, Any]:
+        row = self.db.scalar(
+            select(CharacterCard).where(CharacterCard.user_id == self.user_id).limit(1)
+        )
+        if not row:
+            return normalize_card(DEFAULT_CARD)
+        try:
+            return normalize_card(json.loads(row.card_json or "{}"))
+        except json.JSONDecodeError:
+            return normalize_card(DEFAULT_CARD)
+
+    def save_card(self, raw: dict[str, Any]) -> dict[str, Any]:
+        card = normalize_card(raw)
+        row = self.db.scalar(
+            select(CharacterCard).where(CharacterCard.user_id == self.user_id).limit(1)
+        )
+        if not row:
+            row = CharacterCard(user_id=self.user_id, card_json="{}")
+            self.db.add(row)
+            self.db.flush()
+        row.card_json = json.dumps(card, ensure_ascii=False)
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return card
+
+    def get_system_prompt(self) -> str:
+        return build_system_prompt(self.get_card())

backend/app/chat/generation.py

@@ -0,0 +1,95 @@
+import asyncio
+import logging
+from dataclasses import dataclass, field
+
+from app.chat.service import ChatService
+from app.db.base import SessionLocal
+
+logger = logging.getLogger(__name__)
+
+
+class GenerationBusyError(Exception):
+    """Сессия уже генерирует ответ."""
+
+
+@dataclass
+class GenerationHandle:
+    session_id: int
+    user_id: int
+    user_text: str
+    task: asyncio.Task | None = None
+    subscribers: list[asyncio.Queue[str | None]] = field(default_factory=list)
+    _lock: asyncio.Lock = field(default_factory=asyncio.Lock)
+
+    async def broadcast(self, chunk: str | None) -> None:
+        async with self._lock:
+            targets = list(self.subscribers)
+        for queue in targets:
+            try:
+                queue.put_nowait(chunk)
+            except asyncio.QueueFull:
+                logger.debug("generation queue full for session=%s, dropping subscriber", self.session_id)
+
+    def add_subscriber(self) -> asyncio.Queue[str | None]:
+        queue: asyncio.Queue[str | None] = asyncio.Queue(maxsize=512)
+        self.subscribers.append(queue)
+        return queue
+
+    def remove_subscriber(self, queue: asyncio.Queue[str | None]) -> None:
+        if queue in self.subscribers:
+            self.subscribers.remove(queue)
+
+
+_registry: dict[int, GenerationHandle] = {}
+_registry_lock = asyncio.Lock()
+
+
+def is_generation_active(session_id: int) -> bool:
+    return session_id in _registry
+
+
+def get_active_handle(session_id: int) -> GenerationHandle | None:
+    return _registry.get(session_id)
+
+
+async def _run_generation(handle: GenerationHandle) -> None:
+    db = SessionLocal()
+    try:
+        service = ChatService(db, handle.user_id)
+        async for chunk in service.stream_response(
+            handle.session_id,
+            handle.user_text,
+            user_message_saved=True,
+        ):
+            await handle.broadcast(chunk)
+    except Exception as exc:
+        logger.exception("Background generation failed session=%s", handle.session_id)
+        await handle.broadcast(ChatService._sse("error", {"message": str(exc)}))
+    finally:
+        await handle.broadcast(None)
+        db.close()
+        async with _registry_lock:
+            if _registry.get(handle.session_id) is handle:
+                _registry.pop(handle.session_id, None)
+
+
+async def start_generation(session_id: int, user_id: int, user_text: str) -> GenerationHandle:
+    async with _registry_lock:
+        if session_id in _registry:
+            raise GenerationBusyError()
+        handle = GenerationHandle(session_id=session_id, user_id=user_id, user_text=user_text)
+        _registry[session_id] = handle
+        handle.task = asyncio.create_task(_run_generation(handle))
+        return handle
+
+
+async def subscribe_generation(handle: GenerationHandle):
+    queue = handle.add_subscriber()
+    try:
+        while True:
+            chunk = await queue.get()
+            if chunk is None:
+                break
+            yield chunk
+    finally:
+        handle.remove_subscriber(queue)

backend/app/chat/notice_inbox.py

@@ -1,44 +1,47 @@
-"""Инжект системных оповещений в чат без role=assistant (не ломает LLM-историю)."""
-
-from sqlalchemy import select
-
-from app.db.base import SessionLocal
-from app.db.models import ChatSession, Message
-
-DISPLAY_ONLY_ROLES = frozenset({"notice", "character"})
-
-
-def _latest_chat_session(db) -> ChatSession:
-    session = db.scalar(
-        select(ChatSession).order_by(ChatSession.updated_at.desc()).limit(1)
-    )
-    if not session:
-        session = ChatSession(title="Уведомления")
-        db.add(session)
-        db.commit()
-        db.refresh(session)
-    return session
-
-
-def post_notice_to_latest_chat(content: str) -> int | None:
-    """Сохраняет notice в последний активный чат. Возвращает session_id."""
-    db = SessionLocal()
-    try:
-        session = _latest_chat_session(db)
-        db.add(Message(session_id=session.id, role="notice", content=content))
-        db.commit()
-        return session.id
-    finally:
-        db.close()
-
-
-def post_character_comment_to_latest_chat(content: str) -> int | None:
-    """Реплика персонажа в UI; не попадает в контекст LLM (в отличие от assistant)."""
-    db = SessionLocal()
-    try:
-        session = _latest_chat_session(db)
-        db.add(Message(session_id=session.id, role="character", content=content))
-        db.commit()
-        return session.id
-    finally:
-        db.close()
+"""Инжект системных оповещений в чат без role=assistant (не ломает LLM-историю)."""
+
+from sqlalchemy import select
+
+from app.db.base import SessionLocal
+from app.db.models import ChatSession, Message
+
+DISPLAY_ONLY_ROLES = frozenset({"notice", "character"})
+
+
+def _latest_chat_session(db, user_id: int) -> ChatSession:
+    session = db.scalar(
+        select(ChatSession)
+        .where(ChatSession.user_id == user_id)
+        .order_by(ChatSession.updated_at.desc())
+        .limit(1)
+    )
+    if not session:
+        session = ChatSession(user_id=user_id, title="Уведомления")
+        db.add(session)
+        db.commit()
+        db.refresh(session)
+    return session
+
+
+def post_notice_to_latest_chat(content: str, user_id: int) -> int | None:
+    """Сохраняет notice в последний активный чат пользователя. Возвращает session_id."""
+    db = SessionLocal()
+    try:
+        session = _latest_chat_session(db, user_id)
+        db.add(Message(session_id=session.id, role="notice", content=content))
+        db.commit()
+        return session.id
+    finally:
+        db.close()
+
+
+def post_character_comment_to_latest_chat(content: str, user_id: int) -> int | None:
+    """Реплика персонажа в UI; не попадает в контекст LLM (в отличие от assistant)."""
+    db = SessionLocal()
+    try:
+        session = _latest_chat_session(db, user_id)
+        db.add(Message(session_id=session.id, role="character", content=content))
+        db.commit()
+        return session.id
+    finally:
+        db.close()

backend/app/chat/notices.py

@@ -1,397 +1,432 @@
-import json
-from typing import Any
-
-from app.db.models import PomodoroSession
-from app.pomodoro.cycle import PHASE_LONG_BREAK, PHASE_SHORT_BREAK, PHASE_WORK
-
-PHASE_LABELS = {
-    PHASE_WORK: "Работа",
-    PHASE_SHORT_BREAK: "Короткий перерыв",
-    PHASE_LONG_BREAK: "Длинный перерыв",
-}
-
-
-def _format_time(seconds: int) -> str:
-    minutes, secs = divmod(max(0, seconds), 60)
-    return f"{minutes:02d}:{secs:02d}"
-
-
-def format_phase_completed_notice(
-    session: PomodoroSession,
-    next_phase: str | None,
-) -> str:
-    phase_label = PHASE_LABELS.get(session.phase, session.phase)
-    task = session.task_note or "без описания"
-    lines = [f"⏱ **{phase_label} завершена** · {session.duration_min} мин · _{task}_"]
-
-    if next_phase == PHASE_SHORT_BREAK:
-        lines.append("Дальше: короткий перерыв ☕")
-    elif next_phase == PHASE_LONG_BREAK:
-        lines.append("Дальше: длинный перерыв 🌴 · цикл почти завершён")
-    elif next_phase == PHASE_WORK:
-        lines.append("Дальше: снова работа 💪")
-    else:
-        lines.append("Цикл сброшен. Можно отдохнуть и начать заново.")
-
-    return "\n".join(lines)
-
-
-POMODORO_TOOL_NAMES = frozenset({
-    "get_pomodoro_status",
-    "start_pomodoro",
-    "start_short_break",
-    "start_long_break",
-    "stop_pomodoro",
-    "skip_pomodoro_phase",
-    "reset_pomodoro_cycle",
-    "get_pomodoro_history",
-})
-
-MEMORY_TOOL_NAMES = frozenset({
-    "remember_fact",
-    "recall_memories",
-    "forget_memory",
-    "update_profile",
-    "update_session_summary",
-})
-
-FITNESS_TOOL_NAMES = frozenset({
-    "get_fitness_summary",
-    "get_fitness_history",
-    "set_fitness_profile",
-    "calc_fitness_targets",
-    "log_meal",
-    "log_water",
-    "log_weight",
-    "log_workout",
-    "lookup_food",
-    "lookup_exercise",
-    "set_fitness_reminder",
-})
-
-# Не засорять чат служебными ответами
-REMINDER_TOOL_NAMES = frozenset({
-    "list_reminders",
-    "create_reminder",
-    "update_reminder",
-    "delete_reminder",
-    "complete_reminder",
-})
-
-SHOPPING_TOOL_NAMES = frozenset({
-    "list_shopping_lists",
-    "create_shopping_list",
-    "add_shopping_items",
-    "check_shopping_item",
-    "remove_shopping_item",
-    "delete_shopping_list",
-})
-
-TOOLS_SKIP_CHAT_NOTICE = frozenset({
-    "get_pomodoro_status",
-    "recall_memories",
-    "get_fitness_summary",
-    "get_fitness_history",
-    "lookup_food",
-    "lookup_exercise",
-    "calc_fitness_targets",
-    "get_weather",
-    "get_morning_briefing",
-    "list_shopping_lists",
-    "list_reminders",
-})
-
-
-def format_tool_notice(tool_name: str, raw_result: str) -> str | None:
-    if tool_name in TOOLS_SKIP_CHAT_NOTICE:
-        return None
-
-    try:
-        data = json.loads(raw_result)
-    except json.JSONDecodeError:
-        return None
-
-    if isinstance(data, dict) and "error" in data:
-        if tool_name in POMODORO_TOOL_NAMES:
-            prefix = "⏱"
-        elif tool_name in MEMORY_TOOL_NAMES:
-            prefix = "🧠"
-        elif tool_name in FITNESS_TOOL_NAMES:
-            prefix = "💪"
-        elif tool_name in SHOPPING_TOOL_NAMES:
-            prefix = "🛒"
-        elif tool_name in REMINDER_TOOL_NAMES:
-            prefix = "📅"
-        else:
-            prefix = "📋"
-        return f"{prefix} {data['error']}"
-
-    if tool_name == "reset_pomodoro_cycle":
-        cycle = data.get("cycle", data)
-        return (
-            "⏱ **Цикл помидоро сброшен** · "
-            f"прогресс: {cycle.get('completed_work_sessions', 0)}/"
-            f"{cycle.get('sessions_until_long_break', 4)}"
-        )
-
-    if tool_name in (
-        "get_pomodoro_status",
-        "start_pomodoro",
-        "start_work",
-        "start_short_break",
-        "start_long_break",
-        "stop_pomodoro",
-        "skip_pomodoro_phase",
-    ):
-        return _format_status_notice(data)
-
-    if tool_name == "get_pomodoro_history":
-        return _format_history_notice(data)
-
-    if tool_name == "create_work_item":
-        return _format_work_item_notice(data)
-
-    if tool_name == "list_work_items":
-        return _format_work_items_list_notice(data)
-
-    if tool_name == "list_taiga_tasks":
-        return _format_taiga_tasks_notice(data)
-
-    if tool_name == "sync_taiga_projects":
-        return f"📋 Синхронизировано проектов Taiga: **{len(data)}**"
-
-    if tool_name == "list_taiga_projects":
-        if not isinstance(data, list) or not data:
-            return "📋 Проекты Taiga не найдены. Вызовите sync_taiga_projects."
-        lines = ["📋 **Проекты:**"]
-        for p in data:
-            gitea = f"{p.get('gitea_owner')}/{p.get('gitea_repo')}" if p.get("gitea_configured") else "—"
-            lines.append(f"- `{p.get('slug')}`: {p.get('name')} · Gitea: {gitea}")
-        return "\n".join(lines)
-
-    if tool_name == "remember_fact" and data.get("ok"):
-        action = "обновлено" if data.get("action") == "updated" else "сохранено"
-        return f"🧠 **Память {action}** · #{data.get('memory_id')}: {data.get('content')}"
-
-    if tool_name == "forget_memory" and data.get("ok"):
-        return f"🧠 **Забыто** · #{data.get('memory_id')}: {data.get('forgotten')}"
-
-    if tool_name == "update_profile" and data.get("ok"):
-        profile = data.get("profile") or {}
-        parts = [f"{k}={v}" for k, v in profile.items() if v]
-        return f"🧠 **Профиль обновлён** · {', '.join(parts) or 'пусто'}"
-
-    if tool_name == "update_session_summary" and data.get("ok"):
-        return "🧠 **Сводка чата сохранена**"
-
-    if tool_name == "log_meal" and data.get("ok"):
-        meal = data.get("meal", {})
-        est = "≈" if meal.get("estimated") else ""
-        return (
-            f"💪 **Приём пищи** · {meal.get('description')} · "
-            f"{est}{meal.get('calories', 0):.0f} ккал "
-            f"(Б{meal.get('protein_g', 0):.0f}/Ж{meal.get('fat_g', 0):.0f}/У{meal.get('carbs_g', 0):.0f})"
-        )
-
-    if tool_name == "log_water" and data.get("ok"):
-        w = data.get("water", {})
-        return f"💪 **Вода** +{w.get('amount_ml')} мл"
-
-    if tool_name == "log_weight" and data.get("ok"):
-        m = data.get("metric", {})
-        return f"💪 **Вес** {m.get('weight_kg')} кг"
-
-    if tool_name == "log_workout" and data.get("ok"):
-        wo = data.get("workout", {})
-        return f"💪 **Тренировка** · {wo.get('title')}"
-
-    if tool_name == "set_fitness_profile" and data.get("ok"):
-        p = data.get("profile", {})
-        return (
-            f"💪 **Профиль** · {p.get('calorie_target')} ккал, "
-            f"вода {p.get('water_l')} л"
-        )
-
-    if tool_name == "set_fitness_reminder" and data.get("ok"):
-        r = data.get("reminder", {})
-        state = "вкл" if r.get("enabled") else "выкл"
-        return f"💪 **Напоминание {r.get('kind')}** · {state}"
-
-    if tool_name == "generate_image" and data.get("ok"):
-        url = data.get("url", "")
-        return f"🎨 **Картинка готова**\n\n![image]({url})"
-
-    if tool_name == "create_shopping_list" and data.get("ok"):
-        lst = data.get("list") or {}
-        action = "создан" if data.get("created") else "уже был"
-        return f"🛒 **Список {action}** · «{lst.get('name')}» (#{lst.get('id')})"
-
-    if tool_name == "add_shopping_items" and data.get("ok"):
-        added = data.get("added") or []
-        names = ", ".join(i.get("text", "") for i in added[:5])
-        extra = f" +{len(added) - 5}" if len(added) > 5 else ""
-        return f"🛒 **Добавлено в «{data.get('list_name')}»** · {names}{extra}"
-
-    if tool_name == "check_shopping_item" and data.get("ok"):
-        item = data.get("item") or {}
-        state = "куплено" if item.get("checked") else "снята отметка"
-        return f"🛒 **{state}** · #{item.get('id')} {item.get('text')}"
-
-    if tool_name == "remove_shopping_item" and data.get("ok"):
-        removed = data.get("removed") or {}
-        return f"🛒 **Удалено** · {removed.get('text')}"
-
-    if tool_name == "delete_shopping_list" and data.get("ok"):
-        return f"🛒 **Список удалён** · «{data.get('name')}»"
-
-    if tool_name == "create_reminder" and data.get("ok"):
-        r = data.get("reminder") or {}
-        rec = r.get("recurrence", "none")
-        rec_label = f" · повтор {rec}" if rec and rec != "none" else ""
-        return f"📅 **Напоминание создано** · {r.get('title')} · {r.get('due_at_local')}{rec_label}"
-
-    if tool_name == "update_reminder" and data.get("ok"):
-        r = data.get("reminder") or {}
-        return f"📅 **Напоминание обновлено** · #{r.get('id')} {r.get('title')}"
-
-    if tool_name == "delete_reminder" and data.get("ok"):
-        return f"📅 **Напоминание удалено** · «{data.get('title')}»"
-
-    if tool_name == "complete_reminder" and data.get("ok"):
-        r = data.get("reminder") or {}
-        return f"📅 **Готово** · {r.get('title')}"
-
-    return None
-
-
-def _format_work_item_notice(data: dict[str, Any]) -> str | None:
-    if data.get("error"):
-        return f"📋 {data['error']}"
-    if not data.get("ok"):
-        return None
-    taiga = data.get("taiga", {})
-    gitea = data.get("gitea", {})
-    lines = [
-        "📋 **Создано:**",
-        f"- Taiga: #{taiga.get('ref')} — {taiga.get('subject')}",
-        f"- URL: {taiga.get('url')}",
-    ]
-    if gitea.get("url"):
-        lines.append(f"- Gitea: {gitea.get('url')}")
-    if data.get("branch"):
-        lines.append(f"- Ветка: `{data['branch']}`")
-    subtasks = data.get("subtasks") or []
-    if subtasks:
-        lines.append("**Подзадачи:**")
-        for t in subtasks:
-            lines.append(f"- #{t.get('ref')} {t.get('subject')}")
-    return "\n".join(lines)
-
-
-def _format_work_items_list_notice(data: Any) -> str | None:
-    if not isinstance(data, list) or not data:
-        return "📋 Локальных work items (созданных ассистентом) нет."
-    lines = ["📋 **Work items ассистента:**"]
-    for item in data[:15]:
-        lines.append(
-            f"- [{item.get('status')}] #{item.get('taiga_ref')} {item.get('title')} "
-            f"({item.get('taiga_slug')})"
-        )
-    return "\n".join(lines)
-
-
-def _format_taiga_tasks_notice(data: Any) -> str | None:
-    if not isinstance(data, dict):
-        return None
-    if data.get("error"):
-        return f"📋 {data['error']}"
-
-    blocks = data.get("projects") or []
-    total_stories = data.get("total_stories", 0)
-    total_tasks = data.get("total_tasks", 0)
-
-    if not blocks or (total_stories == 0 and total_tasks == 0):
-        slug = blocks[0].get("slug") if len(blocks) == 1 else None
-        if slug:
-            return f"📋 В `{slug}` нет открытых user stories и tasks в Taiga."
-        return "📋 Открытых задач в Taiga не найдено."
-
-    lines = [f"📋 **Открытые задачи Taiga** (stories: {total_stories}, tasks: {total_tasks}):"]
-    for block in blocks:
-        stories = block.get("stories") or []
-        tasks = block.get("tasks") or []
-        if not stories and not tasks:
-            continue
-        lines.append(f"**{block.get('name')}** (`{block.get('slug')}`):")
-        for s in stories:
-            lines.append(f"- story #{s.get('ref')} {s.get('subject')}")
-        for t in tasks:
-            lines.append(f"- task #{t.get('ref')} {t.get('subject')}")
-    return "\n".join(lines)
-
-
-def _format_status_notice(data: dict[str, Any]) -> str:
-    status = data.get("status", "idle")
-    phase = data.get("phase", PHASE_WORK)
-    phase_label = PHASE_LABELS.get(phase, phase)
-    task = data.get("task_note") or "без описания"
-    remaining = data.get("remaining_seconds", 0)
-    duration = data.get("duration_min", 25)
-    cycle = data.get("cycle", {})
-    cycle_info = ""
-    if cycle:
-        cycle_info = (
-            f" · цикл {cycle.get('completed_work_sessions', 0)}/"
-            f"{cycle.get('sessions_until_long_break', 4)}"
-        )
-
-    if status == "idle":
-        return f"⏱ **Помидоро:** таймер не запущен{cycle_info}."
-
-    if status == "running":
-        return (
-            f"⏱ **{phase_label}** · осталось **{_format_time(remaining)}** "
-            f"из {duration} мин · _{task}_{cycle_info}"
-        )
-
-    if status == "paused":
-        elapsed = data.get("elapsed_seconds", 0)
-        return (
-            f"⏱ **{phase_label} на паузе** · прошло {_format_time(elapsed)} "
-            f"из {duration} мин · _{task}_{cycle_info}"
-        )
-
-    if status == "completed":
-        return f"⏱ **{phase_label} завершена** · {duration} мин · _{task}_"
-
-    if status == "cancelled":
-        return f"⏱ **{phase_label} отменена** · _{task}_"
-
-    return f"⏱ Помидоро: {status}"
-
-
-def _format_history_notice(data: Any) -> str:
-    if not isinstance(data, list) or not data:
-        return "⏱ **История помидоро** пуста."
-
-    lines = ["⏱ **История помидоро:**"]
-    for item in data[:10]:
-        task = item.get("task_note") or "без описания"
-        phase = PHASE_LABELS.get(item.get("phase", ""), item.get("phase", "?"))
-        duration = item.get("duration_min", "?")
-        lines.append(f"- {phase}: {task} ({duration} мин)")
-
-    return "\n".join(lines)
-
-
-def format_pomodoro_context(status: dict[str, Any]) -> str:
-    notice = _format_status_notice(status)
-    cycle = status.get("cycle", {})
-    extra = ""
-    if cycle:
-        extra = (
-            f"\nНастройки цикла: работа {cycle.get('work_duration_min')} мин, "
-            f"перерыв {cycle.get('short_break_min')} мин, "
-            f"длинный {cycle.get('long_break_min')} мин."
-        )
-    return f"[Актуальный статус помидоро]\n{notice}{extra}"
+import json
+from typing import Any
+
+from app.db.models import PomodoroSession
+from app.pomodoro.cycle import PHASE_LONG_BREAK, PHASE_SHORT_BREAK, PHASE_WORK
+
+PHASE_LABELS = {
+    PHASE_WORK: "Работа",
+    PHASE_SHORT_BREAK: "Короткий перерыв",
+    PHASE_LONG_BREAK: "Длинный перерыв",
+}
+
+
+def _format_time(seconds: int) -> str:
+    minutes, secs = divmod(max(0, seconds), 60)
+    return f"{minutes:02d}:{secs:02d}"
+
+
+def format_phase_completed_notice(
+    session: PomodoroSession,
+    next_phase: str | None,
+) -> str:
+    phase_label = PHASE_LABELS.get(session.phase, session.phase)
+    task = session.task_note or "без описания"
+    lines = [f"⏱ **{phase_label} завершена** · {session.duration_min} мин · _{task}_"]
+
+    if next_phase == PHASE_SHORT_BREAK:
+        lines.append("Дальше: короткий перерыв ☕")
+    elif next_phase == PHASE_LONG_BREAK:
+        lines.append("Дальше: длинный перерыв 🌴 · цикл почти завершён")
+    elif next_phase == PHASE_WORK:
+        lines.append("Дальше: снова работа 💪")
+    else:
+        lines.append("Цикл сброшен. Можно отдохнуть и начать заново.")
+
+    return "\n".join(lines)
+
+
+POMODORO_TOOL_NAMES = frozenset({
+    "get_pomodoro_status",
+    "start_pomodoro",
+    "start_short_break",
+    "start_long_break",
+    "stop_pomodoro",
+    "skip_pomodoro_phase",
+    "reset_pomodoro_cycle",
+    "get_pomodoro_history",
+})
+
+MEMORY_TOOL_NAMES = frozenset({
+    "remember_fact",
+    "recall_memories",
+    "forget_memory",
+    "update_profile",
+    "update_session_summary",
+})
+
+FITNESS_TOOL_NAMES = frozenset({
+    "get_fitness_summary",
+    "get_fitness_history",
+    "set_fitness_profile",
+    "calc_fitness_targets",
+    "calc_body_composition",
+    "log_meal",
+    "log_water",
+    "log_weight",
+    "log_workout",
+    "lookup_food",
+    "lookup_exercise",
+    "set_fitness_reminder",
+})
+
+# Не засорять чат служебными ответами
+REMINDER_TOOL_NAMES = frozenset({
+    "list_reminders",
+    "create_reminder",
+    "update_reminder",
+    "delete_reminder",
+    "complete_reminder",
+})
+
+SHOPPING_TOOL_NAMES = frozenset({
+    "list_shopping_lists",
+    "create_shopping_list",
+    "add_shopping_items",
+    "check_shopping_item",
+    "remove_shopping_item",
+    "delete_shopping_list",
+})
+
+TOOLS_SKIP_CHAT_NOTICE = frozenset({
+    "get_pomodoro_status",
+    "recall_memories",
+    "get_fitness_summary",
+    "get_fitness_history",
+    "lookup_food",
+    "lookup_exercise",
+    "calc_fitness_targets",
+    "calc_body_composition",
+    "get_weather",
+    "get_morning_briefing",
+    "list_shopping_lists",
+    "list_reminders",
+})
+
+
+
+def _format_body_composition_notice(computed: dict[str, Any], *, headline: str) -> str:
+    parts: list[str] = []
+    bf = computed.get("body_fat_pct")
+    if bf is not None:
+        method = computed.get("body_fat_method")
+        if method == "navy":
+            parts.append(f"жир ≈{bf}% (Navy)")
+        elif method == "manual":
+            parts.append(f"жир {bf}%")
+        else:
+            parts.append(f"жир ≈{bf}%")
+    if computed.get("whr") is not None:
+        parts.append(f"WHR {computed.get('whr')}")
+    if computed.get("ffmi") is not None:
+        parts.append(f"FFMI {computed.get('ffmi')}")
+    if parts:
+        return f"{headline} — {', '.join(parts)}"
+    return headline
+
+def format_tool_notice(tool_name: str, raw_result: str) -> str | None:
+    if tool_name in TOOLS_SKIP_CHAT_NOTICE:
+        return None
+
+    try:
+        data = json.loads(raw_result)
+    except json.JSONDecodeError:
+        return None
+
+    if isinstance(data, dict) and "error" in data:
+        if tool_name in POMODORO_TOOL_NAMES:
+            prefix = "⏱"
+        elif tool_name in MEMORY_TOOL_NAMES:
+            prefix = "🧠"
+        elif tool_name in FITNESS_TOOL_NAMES:
+            prefix = "💪"
+        elif tool_name in SHOPPING_TOOL_NAMES:
+            prefix = "🛒"
+        elif tool_name in REMINDER_TOOL_NAMES:
+            prefix = "📅"
+        else:
+            prefix = "📋"
+        return f"{prefix} {data['error']}"
+
+    if tool_name == "reset_pomodoro_cycle":
+        cycle = data.get("cycle", data)
+        return (
+            "⏱ **Цикл помидоро сброшен** · "
+            f"прогресс: {cycle.get('completed_work_sessions', 0)}/"
+            f"{cycle.get('sessions_until_long_break', 4)}"
+        )
+
+    if tool_name in (
+        "get_pomodoro_status",
+        "start_pomodoro",
+        "start_work",
+        "start_short_break",
+        "start_long_break",
+        "stop_pomodoro",
+        "skip_pomodoro_phase",
+    ):
+        return _format_status_notice(data)
+
+    if tool_name == "get_pomodoro_history":
+        return _format_history_notice(data)
+
+    if tool_name == "create_work_item":
+        return _format_work_item_notice(data)
+
+    if tool_name == "list_work_items":
+        return _format_work_items_list_notice(data)
+
+    if tool_name == "list_taiga_tasks":
+        return _format_taiga_tasks_notice(data)
+
+    if tool_name == "sync_taiga_projects":
+        return f"📋 Синхронизировано проектов Taiga: **{len(data)}**"
+
+    if tool_name == "list_taiga_projects":
+        if not isinstance(data, list) or not data:
+            return "📋 Проекты Taiga не найдены. Вызовите sync_taiga_projects."
+        lines = ["📋 **Проекты:**"]
+        for p in data:
+            gitea = f"{p.get('gitea_owner')}/{p.get('gitea_repo')}" if p.get("gitea_configured") else "—"
+            lines.append(f"- `{p.get('slug')}`: {p.get('name')} · Gitea: {gitea}")
+        return "\n".join(lines)
+
+    if tool_name == "remember_fact" and data.get("ok"):
+        action = "обновлено" if data.get("action") == "updated" else "сохранено"
+        return f"🧠 **Память {action}** · #{data.get('memory_id')}: {data.get('content')}"
+
+    if tool_name == "forget_memory" and data.get("ok"):
+        return f"🧠 **Забыто** · #{data.get('memory_id')}: {data.get('forgotten')}"
+
+    if tool_name == "update_profile" and data.get("ok"):
+        profile = data.get("profile") or {}
+        parts = [f"{k}={v}" for k, v in profile.items() if v]
+        return f"🧠 **Профиль обновлён** · {', '.join(parts) or 'пусто'}"
+
+    if tool_name == "update_session_summary" and data.get("ok"):
+        return "🧠 **Сводка чата сохранена**"
+
+    if tool_name == "log_meal" and data.get("ok"):
+        meal = data.get("meal", {})
+        est = "≈" if meal.get("estimated") else ""
+        return (
+            f"💪 **Приём пищи** · {meal.get('description')} · "
+            f"{est}{meal.get('calories', 0):.0f} ккал "
+            f"(Б{meal.get('protein_g', 0):.0f}/Ж{meal.get('fat_g', 0):.0f}/У{meal.get('carbs_g', 0):.0f})"
+        )
+
+    if tool_name == "log_water" and data.get("ok"):
+        w = data.get("water", {})
+        return f"💪 **Вода** +{w.get('amount_ml')} мл"
+
+    if tool_name == "log_weight" and data.get("ok"):
+        m = data.get("metric", {})
+        computed = data.get("computed") or {}
+        headline = f"💪 **Вес** {m.get('weight_kg')} кг"
+        return _format_body_composition_notice(computed, headline=headline)
+
+    if tool_name == "calc_body_composition" and isinstance(data, dict) and "error" not in data:
+        w = data.get("weight_kg")
+        headline = "💪 **Состав тела** (расчёт)"
+        if w is not None:
+            headline += f" · {w} кг"
+        msg = _format_body_composition_notice(data, headline=headline)
+        warnings = data.get("warnings") or []
+        if warnings:
+            msg += f" · {'; '.join(warnings[:2])}"
+        return msg
+
+    if tool_name == "log_workout" and data.get("ok"):
+        wo = data.get("workout", {})
+        return f"💪 **Тренировка** · {wo.get('title')}"
+
+    if tool_name == "set_fitness_profile" and data.get("ok"):
+        p = data.get("profile", {})
+        return (
+            f"💪 **Профиль** · {p.get('calorie_target')} ккал, "
+            f"вода {p.get('water_l')} л"
+        )
+
+    if tool_name == "set_fitness_reminder" and data.get("ok"):
+        r = data.get("reminder", {})
+        state = "вкл" if r.get("enabled") else "выкл"
+        return f"💪 **Напоминание {r.get('kind')}** · {state}"
+
+    if tool_name == "generate_image" and data.get("ok"):
+        url = data.get("url", "")
+        return f"🎨 **Картинка готова**\n\n![image]({url})"
+
+    if tool_name == "create_shopping_list" and data.get("ok"):
+        lst = data.get("list") or {}
+        action = "создан" if data.get("created") else "уже был"
+        return f"🛒 **Список {action}** · «{lst.get('name')}» (#{lst.get('id')})"
+
+    if tool_name == "add_shopping_items" and data.get("ok"):
+        added = data.get("added") or []
+        names = ", ".join(i.get("text", "") for i in added[:5])
+        extra = f" +{len(added) - 5}" if len(added) > 5 else ""
+        return f"🛒 **Добавлено в «{data.get('list_name')}»** · {names}{extra}"
+
+    if tool_name == "check_shopping_item" and data.get("ok"):
+        item = data.get("item") or {}
+        state = "куплено" if item.get("checked") else "снята отметка"
+        return f"🛒 **{state}** · #{item.get('id')} {item.get('text')}"
+
+    if tool_name == "remove_shopping_item" and data.get("ok"):
+        removed = data.get("removed") or {}
+        return f"🛒 **Удалено** · {removed.get('text')}"
+
+    if tool_name == "delete_shopping_list" and data.get("ok"):
+        return f"🛒 **Список удалён** · «{data.get('name')}»"
+
+    if tool_name == "create_reminder" and data.get("ok"):
+        r = data.get("reminder") or {}
+        rec = r.get("recurrence", "none")
+        rec_label = f" · повтор {rec}" if rec and rec != "none" else ""
+        return f"📅 **Напоминание создано** · {r.get('title')} · {r.get('due_at_local')}{rec_label}"
+
+    if tool_name == "update_reminder" and data.get("ok"):
+        r = data.get("reminder") or {}
+        return f"📅 **Напоминание обновлено** · #{r.get('id')} {r.get('title')}"
+
+    if tool_name == "delete_reminder" and data.get("ok"):
+        return f"📅 **Напоминание удалено** · «{data.get('title')}»"
+
+    if tool_name == "complete_reminder" and data.get("ok"):
+        r = data.get("reminder") or {}
+        return f"📅 **Готово** · {r.get('title')}"
+
+    return None
+
+
+def _format_work_item_notice(data: dict[str, Any]) -> str | None:
+    if data.get("error"):
+        return f"📋 {data['error']}"
+    if not data.get("ok"):
+        return None
+    taiga = data.get("taiga", {})
+    gitea = data.get("gitea", {})
+    lines = [
+        "📋 **Создано:**",
+        f"- Taiga: #{taiga.get('ref')} — {taiga.get('subject')}",
+        f"- URL: {taiga.get('url')}",
+    ]
+    if gitea.get("url"):
+        lines.append(f"- Gitea: {gitea.get('url')}")
+    if data.get("branch"):
+        lines.append(f"- Ветка: `{data['branch']}`")
+    subtasks = data.get("subtasks") or []
+    if subtasks:
+        lines.append("**Подзадачи:**")
+        for t in subtasks:
+            lines.append(f"- #{t.get('ref')} {t.get('subject')}")
+    return "\n".join(lines)
+
+
+def _format_work_items_list_notice(data: Any) -> str | None:
+    if not isinstance(data, list) or not data:
+        return "📋 Локальных work items (созданных ассистентом) нет."
+    lines = ["📋 **Work items ассистента:**"]
+    for item in data[:15]:
+        lines.append(
+            f"- [{item.get('status')}] #{item.get('taiga_ref')} {item.get('title')} "
+            f"({item.get('taiga_slug')})"
+        )
+    return "\n".join(lines)
+
+
+def _format_taiga_tasks_notice(data: Any) -> str | None:
+    if not isinstance(data, dict):
+        return None
+    if data.get("error"):
+        return f"📋 {data['error']}"
+
+    blocks = data.get("projects") or []
+    total_stories = data.get("total_stories", 0)
+    total_tasks = data.get("total_tasks", 0)
+
+    if not blocks or (total_stories == 0 and total_tasks == 0):
+        slug = blocks[0].get("slug") if len(blocks) == 1 else None
+        if slug:
+            return f"📋 В `{slug}` нет открытых user stories и tasks в Taiga."
+        return "📋 Открытых задач в Taiga не найдено."
+
+    lines = [f"📋 **Открытые задачи Taiga** (stories: {total_stories}, tasks: {total_tasks}):"]
+    for block in blocks:
+        stories = block.get("stories") or []
+        tasks = block.get("tasks") or []
+        if not stories and not tasks:
+            continue
+        lines.append(f"**{block.get('name')}** (`{block.get('slug')}`):")
+        for s in stories:
+            lines.append(f"- story #{s.get('ref')} {s.get('subject')}")
+        for t in tasks:
+            lines.append(f"- task #{t.get('ref')} {t.get('subject')}")
+    return "\n".join(lines)
+
+
+def _format_status_notice(data: dict[str, Any]) -> str:
+    status = data.get("status", "idle")
+    phase = data.get("phase", PHASE_WORK)
+    phase_label = PHASE_LABELS.get(phase, phase)
+    task = data.get("task_note") or "без описания"
+    remaining = data.get("remaining_seconds", 0)
+    duration = data.get("duration_min", 25)
+    cycle = data.get("cycle", {})
+    cycle_info = ""
+    if cycle:
+        cycle_info = (
+            f" · цикл {cycle.get('completed_work_sessions', 0)}/"
+            f"{cycle.get('sessions_until_long_break', 4)}"
+        )
+
+    if status == "idle":
+        return f"⏱ **Помидоро:** таймер не запущен{cycle_info}."
+
+    if status == "running":
+        return (
+            f"⏱ **{phase_label}** · осталось **{_format_time(remaining)}** "
+            f"из {duration} мин · _{task}_{cycle_info}"
+        )
+
+    if status == "paused":
+        elapsed = data.get("elapsed_seconds", 0)
+        return (
+            f"⏱ **{phase_label} на паузе** · прошло {_format_time(elapsed)} "
+            f"из {duration} мин · _{task}_{cycle_info}"
+        )
+
+    if status == "completed":
+        return f"⏱ **{phase_label} завершена** · {duration} мин · _{task}_"
+
+    if status == "cancelled":
+        return f"⏱ **{phase_label} отменена** · _{task}_"
+
+    return f"⏱ Помидоро: {status}"
+
+
+def _format_history_notice(data: Any) -> str:
+    if not isinstance(data, list) or not data:
+        return "⏱ **История помидоро** пуста."
+
+    lines = ["⏱ **История помидоро:**"]
+    for item in data[:10]:
+        task = item.get("task_note") or "без описания"
+        phase = PHASE_LABELS.get(item.get("phase", ""), item.get("phase", "?"))
+        duration = item.get("duration_min", "?")
+        lines.append(f"- {phase}: {task} ({duration} мин)")
+
+    return "\n".join(lines)
+
+
+def format_pomodoro_context(status: dict[str, Any]) -> str:
+    notice = _format_status_notice(status)
+    cycle = status.get("cycle", {})
+    extra = ""
+    if cycle:
+        extra = (
+            f"\nНастройки цикла: работа {cycle.get('work_duration_min')} мин, "
+            f"перерыв {cycle.get('short_break_min')} мин, "
+            f"длинный {cycle.get('long_break_min')} мин."
+        )
+    return f"[Актуальный статус помидоро]\n{notice}{extra}"

backend/app/chat/service.py.refactor_bak

@@ -0,0 +1,468 @@
+import asyncio
+import json
+import logging
+import time
+from collections.abc import AsyncIterator
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.config import get_settings
+from app.db.base import SessionLocal
+from app.character.service import CharacterService
+from app.chat.history import sanitize_openai_messages, strip_historical_reasoning
+from app.chat.notice_inbox import DISPLAY_ONLY_ROLES
+from app.chat.notices import (
+    POMODORO_TOOL_NAMES,
+    format_pomodoro_context,
+    format_tool_notice,
+)
+from app.fitness.context import format_fitness_context, get_fitness_snapshot
+from app.homelab.context import format_datetime_context
+from app.homelab.openmeteo import format_weather_snapshot
+from app.memory.context import (
+    format_identity_hint,
+    format_memory_context,
+    get_memory_snapshot,
+)
+from app.memory.extract import extract_after_turn
+from app.projects.context import format_projects_context, get_projects_snapshot
+from app.reminders.context import format_reminders_context, get_reminders_snapshot
+from app.shopping.context import format_shopping_context, get_shopping_snapshot
+from app.db.models import ChatSession, Message
+from app.llm.client import LLMClient
+from app.pomodoro.service import PomodoroService
+from app.tools.registry import TOOL_DEFINITIONS, execute_tool
+
+MAX_TOOL_ROUNDS = 5
+MAX_HISTORY_MESSAGES = 40
+
+logger = logging.getLogger(__name__)
+
+
+def _build_messages_for_session(session_id: int) -> list[dict[str, Any]]:
+    db = SessionLocal()
+    try:
+        service = ChatService(db)
+        session = service.get_session(session_id)
+        if not session:
+            return []
+        return service._build_messages(session)
+    finally:
+        db.close()
+
+
+async def _extract_memory_background(
+    session_id: int,
+    user_text: str,
+    assistant_text: str,
+) -> None:
+    db = SessionLocal()
+    try:
+        await extract_after_turn(db, session_id, user_text, assistant_text)
+    except Exception as exc:
+        logger.warning("Background memory extraction failed: %s", exc)
+    finally:
+        db.close()
+
+
+class ChatService:
+    def __init__(self, db: Session):
+        self.db = db
+        self.llm = LLMClient()
+        self.character = CharacterService()
+
+    def list_sessions(self) -> list[ChatSession]:
+        stmt = select(ChatSession).order_by(ChatSession.updated_at.desc())
+        return list(self.db.scalars(stmt).all())
+
+    def get_session(self, session_id: int) -> ChatSession | None:
+        return self.db.get(ChatSession, session_id)
+
+    def create_session(self, title: str = "Новый чат") -> ChatSession:
+        session = ChatSession(title=title)
+        self.db.add(session)
+        self.db.commit()
+        self.db.refresh(session)
+        return session
+
+    def delete_session(self, session_id: int) -> bool:
+        session = self.get_session(session_id)
+        if not session:
+            return False
+        self.db.delete(session)
+        self.db.commit()
+        return True
+
+    def _build_system_prompt(self, session_id: int | None = None) -> str:
+        status = PomodoroService(self.db).get_status()
+        memory_snapshot = get_memory_snapshot(self.db, session_id)
+        fitness_snapshot = get_fitness_snapshot(self.db)
+        shopping_snapshot = get_shopping_snapshot(self.db)
+        reminders_snapshot = get_reminders_snapshot(self.db)
+        projects_snapshot = get_projects_snapshot(self.db)
+        return (
+            f"{self.character.get_system_prompt()}\n\n"
+            f"{format_datetime_context(self.db)}\n\n"
+            f"{format_memory_context(memory_snapshot)}\n\n"
+            f"{format_fitness_context(fitness_snapshot)}\n\n"
+            f"{format_shopping_context(shopping_snapshot)}\n\n"
+            f"{format_reminders_context(reminders_snapshot)}\n\n"
+            f"{format_weather_snapshot()}\n\n"
+            f"{format_pomodoro_context(status)}\n\n"
+            f"{format_projects_context(projects_snapshot)}"
+        )
+
+    def _build_messages(self, session: ChatSession) -> list[dict[str, Any]]:
+        system_prompt = self._build_system_prompt(session.id)
+        all_chat = [m for m in session.messages if m.role not in DISPLAY_ONLY_ROLES]
+        last_user = next((m.content for m in reversed(all_chat) if m.role == "user"), "")
+        if last_user:
+            memory_snapshot = get_memory_snapshot(self.db, session.id)
+            identity_hint = format_identity_hint(memory_snapshot, last_user)
+            if identity_hint:
+                system_prompt += f"\n\n{identity_hint}"
+        if len(all_chat) > MAX_HISTORY_MESSAGES:
+            system_prompt += (
+                f"\n\n[История чата: в контексте последние {MAX_HISTORY_MESSAGES} "
+                f"из {len(all_chat)} сообщений. Раннее — в сводке сессии, если сохранена.]"
+            )
+        messages: list[dict[str, Any]] = [
+            {"role": "system", "content": system_prompt}
+        ]
+        chat_messages = all_chat[-MAX_HISTORY_MESSAGES:] if len(all_chat) > MAX_HISTORY_MESSAGES else all_chat
+
+        for msg in chat_messages:
+            content = msg.content or None
+            entry: dict[str, Any] = {"role": msg.role, "content": content}
+            if msg.tool_calls_json:
+                entry["tool_calls"] = json.loads(msg.tool_calls_json)
+                if not content:
+                    entry["content"] = None
+            reasoning_data = LLMClient.deserialize_reasoning(msg.reasoning_json)
+            if reasoning_data:
+                LLMClient.attach_reasoning_to_message(
+                    entry,
+                    reasoning=reasoning_data.get("reasoning", ""),
+                    reasoning_details=reasoning_data.get("reasoning_details"),
+                )
+            if msg.role == "tool" and msg.tool_call_id:
+                entry["tool_call_id"] = msg.tool_call_id
+            messages.append(entry)
+        messages = sanitize_openai_messages(messages)
+        messages = strip_historical_reasoning(messages)
+        return messages
+
+    def _save_message(
+        self,
+        session_id: int,
+        role: str,
+        content: str = "",
+        tool_calls: list[dict[str, Any]] | None = None,
+        tool_call_id: str | None = None,
+        reasoning_json: str | None = None,
+    ) -> Message:
+        message = Message(
+            session_id=session_id,
+            role=role,
+            content=content,
+            tool_calls_json=json.dumps(tool_calls, ensure_ascii=False) if tool_calls else None,
+            reasoning_json=reasoning_json,
+            tool_call_id=tool_call_id,
+        )
+        self.db.add(message)
+        session = self.get_session(session_id)
+        if session and role == "user" and session.title == "Новый чат" and content:
+            session.title = content[:60] + ("..." if len(content) > 60 else "")
+        self.db.commit()
+        self.db.refresh(message)
+        return message
+
+    def save_user_message(self, session_id: int, user_text: str) -> None:
+        self._save_message(session_id, "user", user_text)
+
+    async def _fallback_complete(
+        self,
+        messages: list[dict[str, Any]],
+        session_id: int,
+    ) -> tuple[str, list[str], list[dict[str, Any]]]:
+        """Нестриминговый запасной путь, если stream вернул пустоту."""
+        logger.info("chat session=%s fallback complete", session_id)
+        result: dict[str, Any] = {"content": "", "tool_calls": []}
+        for with_tools in (True, False):
+            result = await self.llm.complete(
+                messages,
+                tools=TOOL_DEFINITIONS if with_tools else None,
+                temperature=0.5,
+                visible_reply=True,
+            )
+            if (result.get("content") or "").strip() or result.get("tool_calls"):
+                break
+
+        tool_calls = result.get("tool_calls") or []
+        content = (result.get("content") or "").strip()
+        notices: list[str] = []
+        pomodoro_events: list[dict[str, Any]] = []
+
+        if tool_calls:
+            assistant_msg: dict[str, Any] = {
+                "role": "assistant",
+                "content": content or None,
+                "tool_calls": tool_calls,
+            }
+            messages.append(assistant_msg)
+            self._save_message(
+                session_id,
+                "assistant",
+                content,
+                tool_calls=tool_calls,
+            )
+            for tool_call in tool_calls:
+                fn = tool_call["function"]
+                args = LLMClient.parse_tool_arguments(fn.get("arguments", ""))
+                tool_result = await execute_tool(
+                    self.db, fn["name"], args, session_id=session_id
+                )
+                messages.append(
+                    {
+                        "role": "tool",
+                        "tool_call_id": tool_call["id"],
+                        "content": tool_result,
+                    }
+                )
+                self._save_message(
+                    session_id,
+                    "tool",
+                    tool_result,
+                    tool_call_id=tool_call["id"],
+                )
+                notice = format_tool_notice(fn["name"], tool_result)
+                if notice:
+                    self._save_message(session_id, "notice", notice)
+                    notices.append(notice)
+                if fn["name"] in POMODORO_TOOL_NAMES:
+                    pomodoro_events.append(
+                        {"name": fn["name"], "result": json.loads(tool_result)}
+                    )
+
+            followup = await self.llm.complete(
+                messages,
+                tools=None,
+                temperature=0.4,
+                visible_reply=True,
+            )
+            return (followup.get("content") or "").strip(), notices, pomodoro_events
+
+        return content, notices, pomodoro_events
+
+    async def stream_response(
+        self,
+        session_id: int,
+        user_text: str,
+        *,
+        user_message_saved: bool = False,
+    ) -> AsyncIterator[str]:
+        session = self.get_session(session_id)
+        if not session:
+            yield self._sse("error", {"message": "Session not found"})
+            return
+
+        if not user_message_saved:
+            self._save_message(session_id, "user", user_text)
+        yield self._sse("status", {"phase": "preparing"})
+        t0 = time.monotonic()
+        messages = await asyncio.to_thread(_build_messages_for_session, session_id)
+        prepare_sec = time.monotonic() - t0
+        if not messages:
+            yield self._sse("error", {"message": "Session not found"})
+            return
+        yield self._sse("status", {"phase": "generating"})
+        streamed_reply_parts: list[str] = []
+        all_tool_notices: list[str] = []
+        tools_executed = 0
+        tool_round = 0
+
+        for _ in range(MAX_TOOL_ROUNDS):
+            tool_round += 1
+            t_round = time.monotonic()
+            content_parts: list[str] = []
+            tool_calls: list[dict[str, Any]] = []
+            reasoning = ""
+            reasoning_details: list[Any] | None = None
+            finish_reason = ""
+
+            # После tool-раунда стримим вживую; до tools — буфер (иначе текст «переписывает» notice).
+            stream_live = tools_executed > 0
+
+            async for event in self.llm.stream_chat(messages, tools=TOOL_DEFINITIONS):
+                if event["type"] == "content":
+                    content_parts.append(event["content"])
+                    if stream_live:
+                        yield self._sse("token", {"content": event["content"]})
+                elif event["type"] == "reasoning":
+                    reasoning = event.get("reasoning", "") or reasoning
+                    if event.get("reasoning_details"):
+                        reasoning_details = event["reasoning_details"]
+                elif event["type"] == "error":
+                    logger.warning(
+                        "chat session=%s llm_error round=%d prepare=%.2fs: %s",
+                        session_id,
+                        tool_round,
+                        prepare_sec,
+                        event.get("content"),
+                    )
+                    yield self._sse("error", {"message": event.get("content", "LLM error")})
+                    return
+                elif event["type"] == "tool_calls":
+                    tool_calls = event["tool_calls"]
+                elif event["type"] == "done":
+                    finish_reason = event.get("finish_reason", "")
+
+            logger.info(
+                "chat session=%s round=%d prepare=%.2fs llm=%.2fs "
+                "content_len=%d tool_calls=%d finish_reason=%s reasoning_len=%d",
+                session_id,
+                tool_round,
+                prepare_sec,
+                time.monotonic() - t_round,
+                len("".join(content_parts)),
+                len(tool_calls),
+                finish_reason,
+                len(reasoning),
+            )
+
+            if tool_calls:
+                round_text = "".join(content_parts)
+                if round_text.strip():
+                    streamed_reply_parts.append(round_text)
+
+                assistant_msg: dict[str, Any] = {
+                    "role": "assistant",
+                    "content": round_text or None,
+                    "tool_calls": tool_calls,
+                }
+                LLMClient.attach_reasoning_to_message(
+                    assistant_msg,
+                    reasoning=reasoning,
+                    reasoning_details=reasoning_details,
+                )
+                reasoning_json = LLMClient.serialize_reasoning(
+                    reasoning=reasoning,
+                    reasoning_details=reasoning_details,
+                )
+                messages.append(assistant_msg)
+                self._save_message(
+                    session_id,
+                    "assistant",
+                    round_text,
+                    tool_calls=tool_calls,
+                    reasoning_json=reasoning_json,
+                )
+
+                round_notices: list[str] = []
+                for tool_call in tool_calls:
+                    fn = tool_call["function"]
+                    args = LLMClient.parse_tool_arguments(fn.get("arguments", ""))
+                    result = await execute_tool(
+                        self.db, fn["name"], args, session_id=session_id
+                    )
+                    tools_executed += 1
+                    tool_message = {
+                        "role": "tool",
+                        "tool_call_id": tool_call["id"],
+                        "content": result,
+                    }
+                    messages.append(tool_message)
+                    self._save_message(session_id, "tool", result, tool_call_id=tool_call["id"])
+
+                    notice = format_tool_notice(fn["name"], result)
+                    if notice:
+                        self._save_message(session_id, "notice", notice)
+                        round_notices.append(notice)
+                        all_tool_notices.append(notice)
+
+                    if fn["name"] in POMODORO_TOOL_NAMES:
+                        yield self._sse(
+                            "pomodoro",
+                            {"name": fn["name"], "result": json.loads(result)},
+                        )
+
+                yield self._sse("status", {"phase": "tools"})
+                for notice in round_notices:
+                    yield self._sse("notice", {"content": notice})
+
+                continue
+
+            if content_parts and not stream_live:
+                for part in content_parts:
+                    yield self._sse("token", {"content": part})
+
+            final_content = "".join(content_parts).strip()
+            if not final_content and streamed_reply_parts and tools_executed == 0:
+                final_content = "".join(streamed_reply_parts).strip()
+            if not final_content and reasoning:
+                final_content = reasoning.strip()
+            if not final_content and tools_executed:
+                retry = await self.llm.complete(
+                    messages,
+                    tools=None,
+                    temperature=0.4,
+                    visible_reply=True,
+                )
+                final_content = (retry.get("content") or "").strip()
+                if final_content:
+                    yield self._sse("token", {"content": final_content})
+            # Notices уже в чате как role=notice — не дублируем в assistant.
+            if not final_content:
+                final_content, fb_notices, fb_pomodoro = await self._fallback_complete(
+                    messages, session_id
+                )
+                if final_content:
+                    yield self._sse("token", {"content": final_content})
+                for notice in fb_notices:
+                    yield self._sse("notice", {"content": notice})
+                for event in fb_pomodoro:
+                    yield self._sse("pomodoro", event)
+
+            if not final_content:
+                logger.warning(
+                    "chat session=%s empty_reply tools=%d rounds=%d finish_reason=%s",
+                    session_id,
+                    tools_executed,
+                    tool_round,
+                    finish_reason,
+                )
+                yield self._sse(
+                    "error",
+                    {
+                        "message": (
+                            "Модель не вернула ответ (finish_reason="
+                            f"{finish_reason or 'unknown'}). "
+                            "Попробуй новый чат или проверь OPENROUTER_MODEL."
+                        ),
+                    },
+                )
+                return
+
+            self._save_message(session_id, "assistant", final_content)
+
+            logger.info(
+                "chat session=%s done tools=%d reply_len=%d total=%.2fs",
+                session_id,
+                tools_executed,
+                len(final_content),
+                time.monotonic() - t0,
+            )
+            yield self._sse("done", {})
+            if get_settings().memory_auto_extract:
+                asyncio.create_task(
+                    _extract_memory_background(session_id, user_text, final_content)
+                )
+            return
+
+        yield self._sse("error", {"message": "Too many tool call rounds"})
+
+    @staticmethod
+    def _sse(event: str, data: dict[str, Any]) -> str:
+        return f"event: {event}\ndata: {json.dumps(data, ensure_ascii=False)}\n\n"

backend/app/config.py

@@ -1,127 +1,138 @@
-from functools import lru_cache
-from pathlib import Path
-
-from pydantic_settings import BaseSettings, SettingsConfigDict
-
-
-class Settings(BaseSettings):
-    model_config = SettingsConfigDict(
-        env_file=(".env", "../.env"),
-        env_file_encoding="utf-8",
-        extra="ignore",
-    )
-
-    host: str = "0.0.0.0"
-    port: int = 8080
-
-    openrouter_api_key: str = ""
-    openrouter_model: str = "deepseek/deepseek-chat"
-    openrouter_base_url: str = "https://openrouter.ai/api/v1"
-    # Отдельная модель для JSON-задач (память, фитнес). Пусто = та же, что OPENROUTER_MODEL.
-    memory_extract_model: str = ""
-    # Некоторые модели (reasoning / без function calling) — выключить tools.
-    openrouter_tools_enabled: bool = True
-    # DeepSeek V4 / reasoning: none | low | medium | high | xhigh. none = без thinking.
-    openrouter_reasoning_effort: str = "none"
-
-    database_url: str = "sqlite:///./data/assistant.db"
-    cors_origins: str = "http://localhost:5173,http://localhost:8080,http://localhost:3000"
-    system_prompt_path: str = "./prompts/assistant.md"
-    memory_auto_extract: bool = True
-
-    # Taiga/Gitea on host (not in Docker) — use host.docker.internal from container
-    taiga_base_url: str = "http://host.docker.internal:9000"
-    taiga_username: str = ""
-    taiga_password: str = ""
-    taiga_public_url: str = "https://taiga.grigowashere.ru"
-
-    gitea_base_url: str = "http://host.docker.internal:3000"
-    gitea_token: str = ""
-    gitea_public_url: str = "https://git.grigowashere.ru"
-    gitea_webhook_secret: str = ""
-
-    repos_dir: str = "/data/repos"
-
-    wger_base_url: str = "https://wger.de/api/v2"
-    openfoodfacts_base_url: str = "https://world.openfoodfacts.org"
-    fitness_reminders_enabled: bool = True
-    reminders_enabled: bool = True
-
-    openmeteo_base_url: str = "http://192.168.1.109:8085"
-    weather_lat: float = 59.9343
-    weather_lon: float = 30.3351
-    weather_location_name: str = "Санкт-Петербург"
-    weather_cache_sec: int = 300
-
-    news_rss_urls: str = (
-        "https://habr.com/ru/rss/all/all/,"
-        "https://www.reddit.com/r/programming/.rss"
-    )
-    news_cache_sec: int = 1800
-    news_max_items: int = 7
-
-    morning_digest_enabled: bool = True
-    morning_digest_hour: int = 8
-    morning_digest_minute: int = 0
-
-    comfyui_base_url: str = "http://192.168.1.109:8188"
-    comfyui_enabled: bool = True
-    # Anima split-model (default): set UNET+CLIP+VAE, leave CHECKPOINT empty
-    comfyui_checkpoint: str = ""
-    comfyui_unet: str = "anima-preview3-base.safetensors"
-    comfyui_clip: str = "qwen_3_06b_base.safetensors"
-    comfyui_vae: str = "qwen_image_vae.safetensors"
-    comfyui_style_lora: str = "anima-preview-3-masterpieces-v5.safetensors"
-    comfyui_style_lora_weight: float = 0.7
-    comfyui_steps: int = 30
-    comfyui_cfg: float = 4.0
-    comfyui_sampler: str = "er_sde"
-    comfyui_scheduler: str = "simple"
-    comfyui_width: int = 1024
-    comfyui_height: int = 720
-    comfyui_negative_prompt: str = (
-        "worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia"
-    )
-    comfyui_poll_interval_sec: float = 2.0
-    comfyui_timeout_sec: float = 180.0
-    comfyui_rofl_enabled: bool = True
-    comfyui_rofl_max_per_day: int = 1
-    comfyui_rofl_probability: float = 0.15
-    comfyui_rofl_min_interval_hours: int = 12
-    generated_media_dir: str = "./data/generated"
-
-    netdata_base_url: str = "http://host.docker.internal:19999"
-    netdata_public_url: str = ""
-    netdata_alerts_enabled: bool = True
-    netdata_poll_interval_sec: int = 120
-
-    rp_chat_base_url: str = "http://host.docker.internal:8201"
-    rp_chat_enabled: bool = True
-    rp_chat_timeout_sec: float = 300.0
-
-    @property
-    def cors_origins_list(self) -> list[str]:
-        return [origin.strip() for origin in self.cors_origins.split(",") if origin.strip()]
-
-    @property
-    def taiga_configured(self) -> bool:
-        return bool(self.taiga_username and self.taiga_password)
-
-    @property
-    def gitea_configured(self) -> bool:
-        return bool(self.gitea_token)
-
-    @property
-    def news_rss_urls_list(self) -> list[str]:
-        return [u.strip() for u in self.news_rss_urls.split(",") if u.strip()]
-
-    def load_system_prompt(self) -> str:
-        path = Path(self.system_prompt_path)
-        if path.is_file():
-            return path.read_text(encoding="utf-8")
-        return "Ты домашний ИИ-ассистент. Общайся на русском."
-
-
-@lru_cache
-def get_settings() -> Settings:
-    return Settings()
+from functools import lru_cache
+from pathlib import Path
+
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    model_config = SettingsConfigDict(
+        env_file=(".env", "../.env"),
+        env_file_encoding="utf-8",
+        extra="ignore",
+    )
+
+    host: str = "0.0.0.0"
+    port: int = 8080
+
+    openrouter_api_key: str = ""
+    openrouter_model: str = "deepseek/deepseek-chat"
+    openrouter_base_url: str = "https://openrouter.ai/api/v1"
+    # Отдельная модель для JSON-задач (память, фитнес). Пусто = та же, что OPENROUTER_MODEL.
+    memory_extract_model: str = ""
+    # Некоторые модели (reasoning / без function calling) — выключить tools.
+    openrouter_tools_enabled: bool = True
+    # DeepSeek V4 / reasoning: none | low | medium | high | xhigh. none = без thinking.
+    openrouter_reasoning_effort: str = "none"
+
+    database_url: str = "sqlite:///./data/assistant.db"
+    cors_origins: str = "http://localhost:5173,http://localhost:8080,http://localhost:3000"
+    system_prompt_path: str = "./prompts/assistant.md"
+    memory_auto_extract: bool = True
+
+    default_user_username: str = "owner"
+    default_user_display_name: str = ""
+    default_api_token: str = ""
+    auth_required: bool = True
+
+    qdrant_url: str = "http://qdrant:6333"
+    embedding_model: str = "openai/text-embedding-3-small"
+    rag_enabled: bool = False
+    rag_top_k: int = 8
+    memory_facts_in_context: int = 8
+
+    # Taiga/Gitea on host (not in Docker) — use host.docker.internal from container
+    taiga_base_url: str = "http://host.docker.internal:9000"
+    taiga_username: str = ""
+    taiga_password: str = ""
+    taiga_public_url: str = "https://taiga.grigowashere.ru"
+
+    gitea_base_url: str = "http://host.docker.internal:3000"
+    gitea_token: str = ""
+    gitea_public_url: str = "https://git.grigowashere.ru"
+    gitea_webhook_secret: str = ""
+
+    repos_dir: str = "/data/repos"
+
+    wger_base_url: str = "https://wger.de/api/v2"
+    openfoodfacts_base_url: str = "https://world.openfoodfacts.org"
+    fitness_reminders_enabled: bool = True
+    reminders_enabled: bool = True
+
+    openmeteo_base_url: str = "http://192.168.1.109:8085"
+    weather_lat: float = 59.9343
+    weather_lon: float = 30.3351
+    weather_location_name: str = "Санкт-Петербург"
+    weather_cache_sec: int = 300
+
+    news_rss_urls: str = (
+        "https://habr.com/ru/rss/all/all/,"
+        "https://www.reddit.com/r/programming/.rss"
+    )
+    news_cache_sec: int = 1800
+    news_max_items: int = 7
+
+    morning_digest_enabled: bool = True
+    morning_digest_hour: int = 8
+    morning_digest_minute: int = 0
+
+    comfyui_base_url: str = "http://192.168.1.109:8188"
+    comfyui_enabled: bool = True
+    # Anima split-model (default): set UNET+CLIP+VAE, leave CHECKPOINT empty
+    comfyui_checkpoint: str = ""
+    comfyui_unet: str = "anima-preview3-base.safetensors"
+    comfyui_clip: str = "qwen_3_06b_base.safetensors"
+    comfyui_vae: str = "qwen_image_vae.safetensors"
+    comfyui_style_lora: str = "anima-preview-3-masterpieces-v5.safetensors"
+    comfyui_style_lora_weight: float = 0.7
+    comfyui_steps: int = 30
+    comfyui_cfg: float = 4.0
+    comfyui_sampler: str = "er_sde"
+    comfyui_scheduler: str = "simple"
+    comfyui_width: int = 1024
+    comfyui_height: int = 720
+    comfyui_negative_prompt: str = (
+        "worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia"
+    )
+    comfyui_poll_interval_sec: float = 2.0
+    comfyui_timeout_sec: float = 180.0
+    comfyui_rofl_enabled: bool = True
+    comfyui_rofl_max_per_day: int = 1
+    comfyui_rofl_probability: float = 0.15
+    comfyui_rofl_min_interval_hours: int = 12
+    generated_media_dir: str = "./data/generated"
+
+    netdata_base_url: str = "http://host.docker.internal:19999"
+    netdata_public_url: str = ""
+    netdata_alerts_enabled: bool = True
+    netdata_poll_interval_sec: int = 120
+
+    rp_chat_base_url: str = "http://host.docker.internal:8201"
+    rp_chat_enabled: bool = True
+    rp_chat_timeout_sec: float = 300.0
+
+    @property
+    def cors_origins_list(self) -> list[str]:
+        return [origin.strip() for origin in self.cors_origins.split(",") if origin.strip()]
+
+    @property
+    def taiga_configured(self) -> bool:
+        return bool(self.taiga_username and self.taiga_password)
+
+    @property
+    def gitea_configured(self) -> bool:
+        return bool(self.gitea_token)
+
+    @property
+    def news_rss_urls_list(self) -> list[str]:
+        return [u.strip() for u in self.news_rss_urls.split(",") if u.strip()]
+
+    def load_system_prompt(self) -> str:
+        path = Path(self.system_prompt_path)
+        if path.is_file():
+            return path.read_text(encoding="utf-8")
+        return "Ты домашний ИИ-ассистент. Общайся на русском."
+
+
+@lru_cache
+def get_settings() -> Settings:
+    return Settings()

backend/app/config.py.refactor_bak

@@ -0,0 +1,127 @@
+from functools import lru_cache
+from pathlib import Path
+
+from pydantic_settings import BaseSettings, SettingsConfigDict
+
+
+class Settings(BaseSettings):
+    model_config = SettingsConfigDict(
+        env_file=(".env", "../.env"),
+        env_file_encoding="utf-8",
+        extra="ignore",
+    )
+
+    host: str = "0.0.0.0"
+    port: int = 8080
+
+    openrouter_api_key: str = ""
+    openrouter_model: str = "deepseek/deepseek-chat"
+    openrouter_base_url: str = "https://openrouter.ai/api/v1"
+    # Отдельная модель для JSON-задач (память, фитнес). Пусто = та же, что OPENROUTER_MODEL.
+    memory_extract_model: str = ""
+    # Некоторые модели (reasoning / без function calling) — выключить tools.
+    openrouter_tools_enabled: bool = True
+    # DeepSeek V4 / reasoning: none | low | medium | high | xhigh. none = без thinking.
+    openrouter_reasoning_effort: str = "none"
+
+    database_url: str = "sqlite:///./data/assistant.db"
+    cors_origins: str = "http://localhost:5173,http://localhost:8080,http://localhost:3000"
+    system_prompt_path: str = "./prompts/assistant.md"
+    memory_auto_extract: bool = True
+
+    # Taiga/Gitea on host (not in Docker) — use host.docker.internal from container
+    taiga_base_url: str = "http://host.docker.internal:9000"
+    taiga_username: str = ""
+    taiga_password: str = ""
+    taiga_public_url: str = "https://taiga.grigowashere.ru"
+
+    gitea_base_url: str = "http://host.docker.internal:3000"
+    gitea_token: str = ""
+    gitea_public_url: str = "https://git.grigowashere.ru"
+    gitea_webhook_secret: str = ""
+
+    repos_dir: str = "/data/repos"
+
+    wger_base_url: str = "https://wger.de/api/v2"
+    openfoodfacts_base_url: str = "https://world.openfoodfacts.org"
+    fitness_reminders_enabled: bool = True
+    reminders_enabled: bool = True
+
+    openmeteo_base_url: str = "http://192.168.1.109:8085"
+    weather_lat: float = 59.9343
+    weather_lon: float = 30.3351
+    weather_location_name: str = "Санкт-Петербург"
+    weather_cache_sec: int = 300
+
+    news_rss_urls: str = (
+        "https://habr.com/ru/rss/all/all/,"
+        "https://www.reddit.com/r/programming/.rss"
+    )
+    news_cache_sec: int = 1800
+    news_max_items: int = 7
+
+    morning_digest_enabled: bool = True
+    morning_digest_hour: int = 8
+    morning_digest_minute: int = 0
+
+    comfyui_base_url: str = "http://192.168.1.109:8188"
+    comfyui_enabled: bool = True
+    # Anima split-model (default): set UNET+CLIP+VAE, leave CHECKPOINT empty
+    comfyui_checkpoint: str = ""
+    comfyui_unet: str = "anima-preview3-base.safetensors"
+    comfyui_clip: str = "qwen_3_06b_base.safetensors"
+    comfyui_vae: str = "qwen_image_vae.safetensors"
+    comfyui_style_lora: str = "anima-preview-3-masterpieces-v5.safetensors"
+    comfyui_style_lora_weight: float = 0.7
+    comfyui_steps: int = 30
+    comfyui_cfg: float = 4.0
+    comfyui_sampler: str = "er_sde"
+    comfyui_scheduler: str = "simple"
+    comfyui_width: int = 1024
+    comfyui_height: int = 720
+    comfyui_negative_prompt: str = (
+        "worst quality, low quality, score_1, score_2, score_3, blurry, jpeg artifacts, sepia"
+    )
+    comfyui_poll_interval_sec: float = 2.0
+    comfyui_timeout_sec: float = 180.0
+    comfyui_rofl_enabled: bool = True
+    comfyui_rofl_max_per_day: int = 1
+    comfyui_rofl_probability: float = 0.15
+    comfyui_rofl_min_interval_hours: int = 12
+    generated_media_dir: str = "./data/generated"
+
+    netdata_base_url: str = "http://host.docker.internal:19999"
+    netdata_public_url: str = ""
+    netdata_alerts_enabled: bool = True
+    netdata_poll_interval_sec: int = 120
+
+    rp_chat_base_url: str = "http://host.docker.internal:8201"
+    rp_chat_enabled: bool = True
+    rp_chat_timeout_sec: float = 300.0
+
+    @property
+    def cors_origins_list(self) -> list[str]:
+        return [origin.strip() for origin in self.cors_origins.split(",") if origin.strip()]
+
+    @property
+    def taiga_configured(self) -> bool:
+        return bool(self.taiga_username and self.taiga_password)
+
+    @property
+    def gitea_configured(self) -> bool:
+        return bool(self.gitea_token)
+
+    @property
+    def news_rss_urls_list(self) -> list[str]:
+        return [u.strip() for u in self.news_rss_urls.split(",") if u.strip()]
+
+    def load_system_prompt(self) -> str:
+        path = Path(self.system_prompt_path)
+        if path.is_file():
+            return path.read_text(encoding="utf-8")
+        return "Ты домашний ИИ-ассистент. Общайся на русском."
+
+
+@lru_cache
+def get_settings() -> Settings:
+    return Settings()

backend/app/db/migrate_fitness.py

@@ -0,0 +1,94 @@
+from sqlalchemy import inspect, text
+
+from app.db.base import engine
+
+
+def _add_column_if_missing(table: str, column: str, ddl: str) -> None:
+    inspector = inspect(engine)
+    if table not in inspector.get_table_names():
+        return
+    columns = {col["name"] for col in inspector.get_columns(table)}
+    if column in columns:
+        return
+    with engine.begin() as conn:
+        conn.execute(text(ddl))
+
+
+def run_fitness_migrations() -> None:
+    inspector = inspect(engine)
+
+    if "fitness_profiles" in inspector.get_table_names():
+        _add_column_if_missing(
+            "fitness_profiles",
+            "baseline_steps",
+            "ALTER TABLE fitness_profiles ADD COLUMN baseline_steps INTEGER",
+        )
+        _add_column_if_missing(
+            "fitness_profiles",
+            "baseline_workout_kcal",
+            "ALTER TABLE fitness_profiles ADD COLUMN baseline_workout_kcal FLOAT",
+        )
+
+    if "workout_logs" in inspector.get_table_names():
+        _add_column_if_missing(
+            "workout_logs",
+            "active_calories",
+            "ALTER TABLE workout_logs ADD COLUMN active_calories FLOAT",
+        )
+        _add_column_if_missing(
+            "workout_logs",
+            "total_calories",
+            "ALTER TABLE workout_logs ADD COLUMN total_calories FLOAT",
+        )
+        _add_column_if_missing(
+            "workout_logs",
+            "steps",
+            "ALTER TABLE workout_logs ADD COLUMN steps INTEGER",
+        )
+
+    if "step_logs" not in inspector.get_table_names():
+        with engine.begin() as conn:
+            conn.execute(
+                text(
+                    "CREATE TABLE step_logs ("
+                    "id INTEGER PRIMARY KEY AUTOINCREMENT, "
+                    "logged_at DATETIME DEFAULT CURRENT_TIMESTAMP, "
+                    "steps INTEGER DEFAULT 0, "
+                    "active_calories FLOAT, "
+                    "source VARCHAR(32) DEFAULT 'manual', "
+                    "notes TEXT DEFAULT ''"
+                    ")"
+                )
+            )
+
+    if "body_metrics" in inspector.get_table_names():
+        _add_column_if_missing(
+            "body_metrics",
+            "neck_cm",
+            "ALTER TABLE body_metrics ADD COLUMN neck_cm FLOAT",
+        )
+        _add_column_if_missing(
+            "body_metrics",
+            "hip_cm",
+            "ALTER TABLE body_metrics ADD COLUMN hip_cm FLOAT",
+        )
+        _add_column_if_missing(
+            "body_metrics",
+            "body_fat_method",
+            "ALTER TABLE body_metrics ADD COLUMN body_fat_method VARCHAR(16)",
+        )
+        _add_column_if_missing(
+            "body_metrics",
+            "whr",
+            "ALTER TABLE body_metrics ADD COLUMN whr FLOAT",
+        )
+        _add_column_if_missing(
+            "body_metrics",
+            "lbm_kg",
+            "ALTER TABLE body_metrics ADD COLUMN lbm_kg FLOAT",
+        )
+        _add_column_if_missing(
+            "body_metrics",
+            "ffmi",
+            "ALTER TABLE body_metrics ADD COLUMN ffmi FLOAT",
+        )

backend/app/db/migrate_multi_user.py

@@ -0,0 +1,249 @@
+from __future__ import annotations
+
+import json
+import logging
+import secrets
+from pathlib import Path
+
+from sqlalchemy import inspect, text
+
+from app.auth.tokens import hash_token
+from app.character.card import DEFAULT_CARD, normalize_card
+from app.config import get_settings
+from app.db.base import engine
+
+logger = logging.getLogger(__name__)
+
+TENANT_TABLES = (
+    "chat_sessions",
+    "user_profile",
+    "memory_facts",
+    "fitness_profiles",
+    "body_metrics",
+    "food_logs",
+    "water_logs",
+    "workout_logs",
+    "step_logs",
+    "fitness_reminders",
+    "shopping_lists",
+    "reminders",
+    "documents",
+    "pomodoro_cycles",
+    "pomodoro_sessions",
+    "project_bindings",
+    "work_items",
+)
+
+LEGACY_CARD_PATH = Path("./data/character.json")
+
+
+def _table_exists(name: str) -> bool:
+    return name in inspect(engine).get_table_names()
+
+
+def _columns(table: str) -> set[str]:
+    if not _table_exists(table):
+        return set()
+    return {col["name"] for col in inspect(engine).get_columns(table)}
+
+
+def _add_column_if_missing(table: str, column: str, ddl: str) -> None:
+    if column in _columns(table):
+        return
+    with engine.begin() as conn:
+        conn.execute(text(ddl))
+
+
+def _ensure_users_table() -> None:
+    if _table_exists("users"):
+        return
+    with engine.begin() as conn:
+        conn.execute(
+            text(
+                "CREATE TABLE users ("
+                "id INTEGER PRIMARY KEY AUTOINCREMENT, "
+                "username VARCHAR(64) NOT NULL UNIQUE, "
+                "display_name VARCHAR(255) DEFAULT '', "
+                "api_token_hash VARCHAR(64) NOT NULL, "
+                "is_active BOOLEAN DEFAULT 1, "
+                "created_at DATETIME DEFAULT CURRENT_TIMESTAMP"
+                ")"
+            )
+        )
+        conn.execute(text("CREATE INDEX IF NOT EXISTS ix_users_api_token_hash ON users (api_token_hash)"))
+
+
+def _ensure_character_cards_table() -> None:
+    if _table_exists("character_cards"):
+        return
+    with engine.begin() as conn:
+        conn.execute(
+            text(
+                "CREATE TABLE character_cards ("
+                "id INTEGER PRIMARY KEY AUTOINCREMENT, "
+                "user_id INTEGER NOT NULL UNIQUE REFERENCES users(id) ON DELETE CASCADE, "
+                "card_json TEXT DEFAULT '{}', "
+                "updated_at DATETIME DEFAULT CURRENT_TIMESTAMP"
+                ")"
+            )
+        )
+        conn.execute(text("CREATE INDEX IF NOT EXISTS ix_character_cards_user_id ON character_cards (user_id)"))
+
+
+def _add_user_id_columns() -> None:
+    for table in TENANT_TABLES:
+        if not _table_exists(table):
+            continue
+        _add_column_if_missing(
+            table,
+            "user_id",
+            f"ALTER TABLE {table} ADD COLUMN user_id INTEGER REFERENCES users(id) ON DELETE CASCADE",
+        )
+        with engine.begin() as conn:
+            conn.execute(text(f"CREATE INDEX IF NOT EXISTS ix_{table}_user_id ON {table} (user_id)"))
+
+
+def _ensure_default_user() -> tuple[int, str | None]:
+    settings = get_settings()
+    with engine.begin() as conn:
+        row = conn.execute(text("SELECT id FROM users ORDER BY id LIMIT 1")).fetchone()
+        if row:
+            return int(row[0]), None
+
+        username = settings.default_user_username or "owner"
+        display_name = settings.default_user_display_name or username
+        plain_token = (settings.default_api_token or "").strip()
+        generated = False
+        if not plain_token:
+            plain_token = secrets.token_urlsafe(32)
+            generated = True
+
+        token_hash = hash_token(plain_token)
+        conn.execute(
+            text(
+                "INSERT INTO users (id, username, display_name, api_token_hash, is_active) "
+                "VALUES (1, :username, :display_name, :token_hash, 1)"
+            ),
+            {"username": username, "display_name": display_name, "token_hash": token_hash},
+        )
+    if generated:
+        logger.warning(
+            "DEFAULT_API_TOKEN not set — generated token for user '%s': %s",
+            username,
+            plain_token,
+        )
+        return 1, plain_token
+    return 1, None
+
+
+def _backfill_user_id(default_user_id: int = 1) -> None:
+    with engine.begin() as conn:
+        for table in TENANT_TABLES:
+            if not _table_exists(table):
+                continue
+            conn.execute(
+                text(f"UPDATE {table} SET user_id = :uid WHERE user_id IS NULL"),
+                {"uid": default_user_id},
+            )
+
+
+def _rebuild_shopping_unique() -> None:
+    if not _table_exists("shopping_lists"):
+        return
+    with engine.begin() as conn:
+        conn.execute(text("CREATE UNIQUE INDEX IF NOT EXISTS uq_shopping_lists_user_name ON shopping_lists (user_id, name)"))
+
+
+def _rebuild_project_bindings_unique() -> None:
+    if not _table_exists("project_bindings"):
+        return
+    with engine.begin() as conn:
+        conn.execute(
+            text(
+                "CREATE UNIQUE INDEX IF NOT EXISTS uq_project_bindings_user_slug "
+                "ON project_bindings (user_id, taiga_slug)"
+            )
+        )
+
+
+def _import_character_card(user_id: int) -> None:
+    with engine.begin() as conn:
+        existing = conn.execute(
+            text("SELECT id FROM character_cards WHERE user_id = :uid"),
+            {"uid": user_id},
+        ).fetchone()
+        if existing:
+            return
+
+        card = normalize_card(DEFAULT_CARD)
+        if LEGACY_CARD_PATH.is_file():
+            try:
+                raw = json.loads(LEGACY_CARD_PATH.read_text(encoding="utf-8"))
+                card = normalize_card(raw)
+            except (json.JSONDecodeError, OSError):
+                pass
+
+        conn.execute(
+            text("INSERT INTO character_cards (user_id, card_json) VALUES (:uid, :json)"),
+            {"uid": user_id, "json": json.dumps(card, ensure_ascii=False)},
+        )
+
+
+def _backfill_qdrant_user_id(default_user_id: int = 1) -> None:
+    settings = get_settings()
+    if not settings.rag_enabled:
+        return
+    try:
+        from app.rag.store import COLLECTION_DOC_CHUNKS, COLLECTION_FACTS, COLLECTION_SUMMARIES, _client
+    except Exception:
+        logger.exception("Qdrant backfill skipped")
+        return
+
+    try:
+        client = _client()
+    except Exception:
+        logger.warning('Qdrant unavailable, skipping user_id backfill')
+        return
+
+    for collection in (COLLECTION_FACTS, COLLECTION_SUMMARIES, COLLECTION_DOC_CHUNKS):
+        try:
+            if not client.collection_exists(collection):
+                continue
+        except Exception:
+            logger.warning('Qdrant unavailable for collection %s', collection)
+            continue
+        offset = None
+        while True:
+            points, offset = client.scroll(
+                collection_name=collection,
+                limit=100,
+                offset=offset,
+                with_payload=True,
+                with_vectors=False,
+            )
+            if not points:
+                break
+            missing = [point.id for point in points if (point.payload or {}).get("user_id") is None]
+            if missing:
+                client.set_payload(
+                    collection_name=collection,
+                    payload={"user_id": default_user_id},
+                    points=missing,
+                )
+            if offset is None:
+                break
+    logger.info("Qdrant user_id backfill completed for user_id=%s", default_user_id)
+
+
+def run_multi_user_migrations() -> str | None:
+    """Returns newly generated API token if any."""
+    _ensure_users_table()
+    _ensure_character_cards_table()
+    _add_user_id_columns()
+    user_id, new_token = _ensure_default_user()
+    _backfill_user_id(user_id)
+    _rebuild_shopping_unique()
+    _rebuild_project_bindings_unique()
+    _import_character_card(user_id)
+    _backfill_qdrant_user_id(user_id)
+    return new_token

backend/app/db/models.py

@@ -1,299 +1,396 @@
-from datetime import datetime
-
-from sqlalchemy import Boolean, DateTime, Float, ForeignKey, Integer, String, Text, func
-from sqlalchemy.orm import Mapped, mapped_column, relationship
-
-from app.db.base import Base
-
-
-class ChatSession(Base):
-    __tablename__ = "chat_sessions"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    title: Mapped[str] = mapped_column(String(255), default="Новый чат")
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-    messages: Mapped[list["Message"]] = relationship(
-        back_populates="session", cascade="all, delete-orphan", order_by="Message.created_at"
-    )
-
-
-class Message(Base):
-    __tablename__ = "messages"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    session_id: Mapped[int] = mapped_column(ForeignKey("chat_sessions.id", ondelete="CASCADE"), index=True)
-    role: Mapped[str] = mapped_column(String(32))
-    content: Mapped[str] = mapped_column(Text, default="")
-    tool_calls_json: Mapped[str | None] = mapped_column(Text, nullable=True)
-    reasoning_json: Mapped[str | None] = mapped_column(Text, nullable=True)
-    tool_call_id: Mapped[str | None] = mapped_column(String(64), nullable=True)
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-
-    session: Mapped["ChatSession"] = relationship(back_populates="messages")
-
-
-class PomodoroCycle(Base):
-    __tablename__ = "pomodoro_cycles"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    work_duration_min: Mapped[int] = mapped_column(Integer, default=25)
-    short_break_min: Mapped[int] = mapped_column(Integer, default=5)
-    long_break_min: Mapped[int] = mapped_column(Integer, default=15)
-    sessions_until_long_break: Mapped[int] = mapped_column(Integer, default=4)
-    completed_work_sessions: Mapped[int] = mapped_column(Integer, default=0)
-    task_note: Mapped[str] = mapped_column(Text, default="")
-    auto_advance: Mapped[bool] = mapped_column(Boolean, default=True)
-    chat_notify_seq: Mapped[int] = mapped_column(Integer, default=0)
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class PomodoroSession(Base):
-    __tablename__ = "pomodoro_sessions"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    status: Mapped[str] = mapped_column(String(32), default="idle")
-    phase: Mapped[str] = mapped_column(String(32), default="work")
-    duration_min: Mapped[int] = mapped_column(Integer, default=25)
-    task_note: Mapped[str] = mapped_column(Text, default="")
-    result: Mapped[str | None] = mapped_column(Text, nullable=True)
-    completed: Mapped[bool] = mapped_column(Boolean, default=False)
-    completion_notified: Mapped[bool] = mapped_column(Boolean, default=False)
-    started_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
-    paused_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
-    elapsed_seconds: Mapped[int] = mapped_column(Integer, default=0)
-    finished_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-
-
-class TaigaProject(Base):
-    __tablename__ = "taiga_projects"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    taiga_id: Mapped[int] = mapped_column(Integer, unique=True, index=True)
-    name: Mapped[str] = mapped_column(String(255))
-    slug: Mapped[str] = mapped_column(String(255), unique=True, index=True)
-    synced_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-
-
-class ProjectBinding(Base):
-    __tablename__ = "project_bindings"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    taiga_slug: Mapped[str] = mapped_column(String(255), unique=True, index=True)
-    gitea_owner: Mapped[str] = mapped_column(String(255), default="")
-    gitea_repo: Mapped[str] = mapped_column(String(255), default="")
-    default_branch: Mapped[str] = mapped_column(String(64), default="main")
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class UserProfile(Base):
-    __tablename__ = "user_profile"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    data_json: Mapped[str] = mapped_column(Text, default="{}")
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class MemoryFact(Base):
-    __tablename__ = "memory_facts"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    category: Mapped[str] = mapped_column(String(64), default="fact", index=True)
-    content: Mapped[str] = mapped_column(Text)
-    source: Mapped[str] = mapped_column(String(32), default="user")
-    session_id: Mapped[int | None] = mapped_column(
-        ForeignKey("chat_sessions.id", ondelete="SET NULL"), nullable=True, index=True
-    )
-    importance: Mapped[int] = mapped_column(Integer, default=3)
-    active: Mapped[bool] = mapped_column(Boolean, default=True, index=True)
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class SessionSummary(Base):
-    __tablename__ = "session_summaries"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    session_id: Mapped[int] = mapped_column(
-        ForeignKey("chat_sessions.id", ondelete="CASCADE"), unique=True, index=True
-    )
-    summary: Mapped[str] = mapped_column(Text, default="")
-    message_count: Mapped[int] = mapped_column(Integer, default=0)
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class FitnessProfile(Base):
-    __tablename__ = "fitness_profiles"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    sex: Mapped[str] = mapped_column(String(16), default="male")
-    age: Mapped[int] = mapped_column(Integer, default=30)
-    height_cm: Mapped[float] = mapped_column(Float, default=170.0)
-    weight_kg: Mapped[float] = mapped_column(Float, default=70.0)
-    activity_level: Mapped[str] = mapped_column(String(32), default="moderate")
-    goal: Mapped[str] = mapped_column(String(32), default="maintain")
-    target_weight_kg: Mapped[float | None] = mapped_column(Float, nullable=True)
-    weekly_workouts: Mapped[int] = mapped_column(Integer, default=3)
-    calorie_target: Mapped[float] = mapped_column(Float, default=2000.0)
-    protein_g: Mapped[float] = mapped_column(Float, default=140.0)
-    fat_g: Mapped[float] = mapped_column(Float, default=65.0)
-    carbs_g: Mapped[float] = mapped_column(Float, default=200.0)
-    water_l: Mapped[float] = mapped_column(Float, default=2.5)
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class BodyMetric(Base):
-    __tablename__ = "body_metrics"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    recorded_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    weight_kg: Mapped[float] = mapped_column(Float)
-    body_fat_pct: Mapped[float | None] = mapped_column(Float, nullable=True)
-    chest_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
-    waist_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
-    notes: Mapped[str] = mapped_column(Text, default="")
-
-
-class FoodLog(Base):
-    __tablename__ = "food_logs"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    meal_type: Mapped[str] = mapped_column(String(32), default="snack")
-    description: Mapped[str] = mapped_column(Text, default="")
-    calories: Mapped[float] = mapped_column(Float, default=0)
-    protein_g: Mapped[float] = mapped_column(Float, default=0)
-    fat_g: Mapped[float] = mapped_column(Float, default=0)
-    carbs_g: Mapped[float] = mapped_column(Float, default=0)
-    source: Mapped[str] = mapped_column(String(32), default="llm")
-    estimated: Mapped[bool] = mapped_column(Boolean, default=True)
-
-
-class WaterLog(Base):
-    __tablename__ = "water_logs"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    amount_ml: Mapped[int] = mapped_column(Integer)
-
-
-class WorkoutLog(Base):
-    __tablename__ = "workout_logs"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    title: Mapped[str] = mapped_column(String(255), default="Тренировка")
-    notes: Mapped[str] = mapped_column(Text, default="")
-    duration_min: Mapped[int | None] = mapped_column(Integer, nullable=True)
-    exercises_json: Mapped[str] = mapped_column(Text, default="[]")
-
-
-class FitnessReminder(Base):
-    __tablename__ = "fitness_reminders"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    kind: Mapped[str] = mapped_column(String(32))
-    hour: Mapped[int] = mapped_column(Integer, default=12)
-    minute: Mapped[int] = mapped_column(Integer, default=0)
-    interval_hours: Mapped[int | None] = mapped_column(Integer, nullable=True)
-    enabled: Mapped[bool] = mapped_column(Boolean, default=True)
-    last_fired_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
-
-
-class ShoppingList(Base):
-    __tablename__ = "shopping_lists"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    name: Mapped[str] = mapped_column(String(255), unique=True, index=True)
-    sort_order: Mapped[int] = mapped_column(Integer, default=0)
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-    items: Mapped[list["ShoppingListItem"]] = relationship(
-        back_populates="shopping_list",
-        cascade="all, delete-orphan",
-        order_by="ShoppingListItem.sort_order, ShoppingListItem.id",
-    )
-
-
-class ShoppingListItem(Base):
-    __tablename__ = "shopping_list_items"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    list_id: Mapped[int] = mapped_column(
-        ForeignKey("shopping_lists.id", ondelete="CASCADE"), index=True
-    )
-    text: Mapped[str] = mapped_column(String(500))
-    quantity: Mapped[float | None] = mapped_column(Float, nullable=True)
-    unit: Mapped[str] = mapped_column(String(64), default="")
-    checked: Mapped[bool] = mapped_column(Boolean, default=False, index=True)
-    sort_order: Mapped[int] = mapped_column(Integer, default=0)
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-
-    shopping_list: Mapped["ShoppingList"] = relationship(back_populates="items")
-
-
-class Reminder(Base):
-    __tablename__ = "reminders"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    title: Mapped[str] = mapped_column(String(255))
-    notes: Mapped[str] = mapped_column(Text, default="")
-    due_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), index=True)
-    all_day: Mapped[bool] = mapped_column(Boolean, default=False)
-    recurrence: Mapped[str] = mapped_column(String(16), default="none")
-    enabled: Mapped[bool] = mapped_column(Boolean, default=True, index=True)
-    last_fired_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
-    completed_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
-    timezone: Mapped[str] = mapped_column(String(64), default="Europe/Moscow")
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class AssistantState(Base):
-    __tablename__ = "assistant_state"
-
-    key: Mapped[str] = mapped_column(String(128), primary_key=True)
-    value: Mapped[str] = mapped_column(Text, default="")
-    updated_at: Mapped[datetime] = mapped_column(
-        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
-    )
-
-
-class WorkItem(Base):
-    __tablename__ = "work_items"
-
-    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
-    taiga_slug: Mapped[str] = mapped_column(String(255), index=True)
-    taiga_project_id: Mapped[int] = mapped_column(Integer)
-    taiga_story_id: Mapped[int] = mapped_column(Integer)
-    taiga_story_ref: Mapped[int] = mapped_column(Integer, index=True)
-    gitea_owner: Mapped[str] = mapped_column(String(255), default="")
-    gitea_repo: Mapped[str] = mapped_column(String(255), default="")
-    gitea_issue_number: Mapped[int | None] = mapped_column(Integer, nullable=True, index=True)
-    suggested_branch: Mapped[str] = mapped_column(String(255), default="")
-    raw_text: Mapped[str] = mapped_column(Text, default="")
-    title: Mapped[str] = mapped_column(String(500), default="")
-    status: Mapped[str] = mapped_column(String(32), default="open")
-    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
-    closed_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+from datetime import datetime
+
+from sqlalchemy import Boolean, DateTime, Float, ForeignKey, Integer, String, Text, UniqueConstraint, func
+from sqlalchemy.orm import Mapped, mapped_column, relationship
+
+from app.db.base import Base
+
+
+class User(Base):
+    __tablename__ = "users"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    username: Mapped[str] = mapped_column(String(64), unique=True, index=True)
+    display_name: Mapped[str] = mapped_column(String(255), default="")
+    api_token_hash: Mapped[str] = mapped_column(String(64), index=True)
+    is_active: Mapped[bool] = mapped_column(Boolean, default=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+
+class CharacterCard(Base):
+    __tablename__ = "character_cards"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(
+        ForeignKey("users.id", ondelete="CASCADE"), unique=True, index=True
+    )
+    card_json: Mapped[str] = mapped_column(Text, default="{}")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class ChatSession(Base):
+    __tablename__ = "chat_sessions"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    title: Mapped[str] = mapped_column(String(255), default="Новый чат")
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+    messages: Mapped[list["Message"]] = relationship(
+        back_populates="session", cascade="all, delete-orphan", order_by="Message.created_at"
+    )
+
+
+class Message(Base):
+    __tablename__ = "messages"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    session_id: Mapped[int] = mapped_column(ForeignKey("chat_sessions.id", ondelete="CASCADE"), index=True)
+    role: Mapped[str] = mapped_column(String(32))
+    content: Mapped[str] = mapped_column(Text, default="")
+    tool_calls_json: Mapped[str | None] = mapped_column(Text, nullable=True)
+    reasoning_json: Mapped[str | None] = mapped_column(Text, nullable=True)
+    tool_call_id: Mapped[str | None] = mapped_column(String(64), nullable=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+    session: Mapped["ChatSession"] = relationship(back_populates="messages")
+
+
+class PomodoroCycle(Base):
+    __tablename__ = "pomodoro_cycles"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    work_duration_min: Mapped[int] = mapped_column(Integer, default=25)
+    short_break_min: Mapped[int] = mapped_column(Integer, default=5)
+    long_break_min: Mapped[int] = mapped_column(Integer, default=15)
+    sessions_until_long_break: Mapped[int] = mapped_column(Integer, default=4)
+    completed_work_sessions: Mapped[int] = mapped_column(Integer, default=0)
+    task_note: Mapped[str] = mapped_column(Text, default="")
+    auto_advance: Mapped[bool] = mapped_column(Boolean, default=True)
+    chat_notify_seq: Mapped[int] = mapped_column(Integer, default=0)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class PomodoroSession(Base):
+    __tablename__ = "pomodoro_sessions"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    status: Mapped[str] = mapped_column(String(32), default="idle")
+    phase: Mapped[str] = mapped_column(String(32), default="work")
+    duration_min: Mapped[int] = mapped_column(Integer, default=25)
+    task_note: Mapped[str] = mapped_column(Text, default="")
+    result: Mapped[str | None] = mapped_column(Text, nullable=True)
+    completed: Mapped[bool] = mapped_column(Boolean, default=False)
+    completion_notified: Mapped[bool] = mapped_column(Boolean, default=False)
+    started_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    paused_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    elapsed_seconds: Mapped[int] = mapped_column(Integer, default=0)
+    finished_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+
+class TaigaProject(Base):
+    __tablename__ = "taiga_projects"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    taiga_id: Mapped[int] = mapped_column(Integer, unique=True, index=True)
+    name: Mapped[str] = mapped_column(String(255))
+    slug: Mapped[str] = mapped_column(String(255), unique=True, index=True)
+    synced_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+
+class ProjectBinding(Base):
+    __tablename__ = "project_bindings"
+    __table_args__ = (UniqueConstraint("user_id", "taiga_slug", name="uq_project_bindings_user_slug"),)
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    taiga_slug: Mapped[str] = mapped_column(String(255), index=True)
+    gitea_owner: Mapped[str] = mapped_column(String(255), default="")
+    gitea_repo: Mapped[str] = mapped_column(String(255), default="")
+    default_branch: Mapped[str] = mapped_column(String(64), default="main")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class UserProfile(Base):
+    __tablename__ = "user_profile"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    data_json: Mapped[str] = mapped_column(Text, default="{}")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class MemoryFact(Base):
+    __tablename__ = "memory_facts"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    category: Mapped[str] = mapped_column(String(64), default="fact", index=True)
+    content: Mapped[str] = mapped_column(Text)
+    source: Mapped[str] = mapped_column(String(32), default="user")
+    session_id: Mapped[int | None] = mapped_column(
+        ForeignKey("chat_sessions.id", ondelete="SET NULL"), nullable=True, index=True
+    )
+    importance: Mapped[int] = mapped_column(Integer, default=3)
+    active: Mapped[bool] = mapped_column(Boolean, default=True, index=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class SessionSummary(Base):
+    __tablename__ = "session_summaries"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    session_id: Mapped[int] = mapped_column(
+        ForeignKey("chat_sessions.id", ondelete="CASCADE"), unique=True, index=True
+    )
+    summary: Mapped[str] = mapped_column(Text, default="")
+    message_count: Mapped[int] = mapped_column(Integer, default=0)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class FitnessProfile(Base):
+    __tablename__ = "fitness_profiles"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    sex: Mapped[str] = mapped_column(String(16), default="male")
+    age: Mapped[int] = mapped_column(Integer, default=30)
+    height_cm: Mapped[float] = mapped_column(Float, default=170.0)
+    weight_kg: Mapped[float] = mapped_column(Float, default=70.0)
+    activity_level: Mapped[str] = mapped_column(String(32), default="moderate")
+    goal: Mapped[str] = mapped_column(String(32), default="maintain")
+    target_weight_kg: Mapped[float | None] = mapped_column(Float, nullable=True)
+    weekly_workouts: Mapped[int] = mapped_column(Integer, default=3)
+    baseline_steps: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    baseline_workout_kcal: Mapped[float | None] = mapped_column(Float, nullable=True)
+    calorie_target: Mapped[float] = mapped_column(Float, default=2000.0)
+    protein_g: Mapped[float] = mapped_column(Float, default=140.0)
+    fat_g: Mapped[float] = mapped_column(Float, default=65.0)
+    carbs_g: Mapped[float] = mapped_column(Float, default=200.0)
+    water_l: Mapped[float] = mapped_column(Float, default=2.5)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class BodyMetric(Base):
+    __tablename__ = "body_metrics"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    recorded_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    weight_kg: Mapped[float] = mapped_column(Float)
+    body_fat_pct: Mapped[float | None] = mapped_column(Float, nullable=True)
+    body_fat_method: Mapped[str | None] = mapped_column(String(16), nullable=True)
+    chest_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
+    waist_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
+    neck_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
+    hip_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
+    whr: Mapped[float | None] = mapped_column(Float, nullable=True)
+    lbm_kg: Mapped[float | None] = mapped_column(Float, nullable=True)
+    ffmi: Mapped[float | None] = mapped_column(Float, nullable=True)
+    notes: Mapped[str] = mapped_column(Text, default="")
+
+
+class FoodLog(Base):
+    __tablename__ = "food_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    meal_type: Mapped[str] = mapped_column(String(32), default="snack")
+    description: Mapped[str] = mapped_column(Text, default="")
+    calories: Mapped[float] = mapped_column(Float, default=0)
+    protein_g: Mapped[float] = mapped_column(Float, default=0)
+    fat_g: Mapped[float] = mapped_column(Float, default=0)
+    carbs_g: Mapped[float] = mapped_column(Float, default=0)
+    source: Mapped[str] = mapped_column(String(32), default="llm")
+    estimated: Mapped[bool] = mapped_column(Boolean, default=True)
+
+
+class StepLog(Base):
+    __tablename__ = "step_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    steps: Mapped[int] = mapped_column(Integer, default=0)
+    active_calories: Mapped[float | None] = mapped_column(Float, nullable=True)
+    source: Mapped[str] = mapped_column(String(32), default="manual")
+    notes: Mapped[str] = mapped_column(Text, default="")
+
+
+class WaterLog(Base):
+    __tablename__ = "water_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    amount_ml: Mapped[int] = mapped_column(Integer)
+
+
+class WorkoutLog(Base):
+    __tablename__ = "workout_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    title: Mapped[str] = mapped_column(String(255), default="Тренировка")
+    notes: Mapped[str] = mapped_column(Text, default="")
+    duration_min: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    active_calories: Mapped[float | None] = mapped_column(Float, nullable=True)
+    total_calories: Mapped[float | None] = mapped_column(Float, nullable=True)
+    steps: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    exercises_json: Mapped[str] = mapped_column(Text, default="[]")
+
+
+class FitnessReminder(Base):
+    __tablename__ = "fitness_reminders"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    kind: Mapped[str] = mapped_column(String(32))
+    hour: Mapped[int] = mapped_column(Integer, default=12)
+    minute: Mapped[int] = mapped_column(Integer, default=0)
+    interval_hours: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    enabled: Mapped[bool] = mapped_column(Boolean, default=True)
+    last_fired_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+
+
+class ShoppingList(Base):
+    __tablename__ = "shopping_lists"
+    __table_args__ = (UniqueConstraint("user_id", "name", name="uq_shopping_lists_user_name"),)
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    name: Mapped[str] = mapped_column(String(255), index=True)
+    sort_order: Mapped[int] = mapped_column(Integer, default=0)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+    items: Mapped[list["ShoppingListItem"]] = relationship(
+        back_populates="shopping_list",
+        cascade="all, delete-orphan",
+        order_by="ShoppingListItem.sort_order, ShoppingListItem.id",
+    )
+
+
+class ShoppingListItem(Base):
+    __tablename__ = "shopping_list_items"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    list_id: Mapped[int] = mapped_column(
+        ForeignKey("shopping_lists.id", ondelete="CASCADE"), index=True
+    )
+    text: Mapped[str] = mapped_column(String(500))
+    quantity: Mapped[float | None] = mapped_column(Float, nullable=True)
+    unit: Mapped[str] = mapped_column(String(64), default="")
+    checked: Mapped[bool] = mapped_column(Boolean, default=False, index=True)
+    sort_order: Mapped[int] = mapped_column(Integer, default=0)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+    shopping_list: Mapped["ShoppingList"] = relationship(back_populates="items")
+
+
+class Reminder(Base):
+    __tablename__ = "reminders"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    title: Mapped[str] = mapped_column(String(255))
+    notes: Mapped[str] = mapped_column(Text, default="")
+    due_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), index=True)
+    all_day: Mapped[bool] = mapped_column(Boolean, default=False)
+    recurrence: Mapped[str] = mapped_column(String(16), default="none")
+    enabled: Mapped[bool] = mapped_column(Boolean, default=True, index=True)
+    last_fired_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    completed_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    timezone: Mapped[str] = mapped_column(String(64), default="Europe/Moscow")
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class AssistantState(Base):
+    __tablename__ = "assistant_state"
+
+    key: Mapped[str] = mapped_column(String(128), primary_key=True)
+    value: Mapped[str] = mapped_column(Text, default="")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class WorkItem(Base):
+    __tablename__ = "work_items"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    taiga_slug: Mapped[str] = mapped_column(String(255), index=True)
+    taiga_project_id: Mapped[int] = mapped_column(Integer)
+    taiga_story_id: Mapped[int] = mapped_column(Integer)
+    taiga_story_ref: Mapped[int] = mapped_column(Integer, index=True)
+    gitea_owner: Mapped[str] = mapped_column(String(255), default="")
+    gitea_repo: Mapped[str] = mapped_column(String(255), default="")
+    gitea_issue_number: Mapped[int | None] = mapped_column(Integer, nullable=True, index=True)
+    suggested_branch: Mapped[str] = mapped_column(String(255), default="")
+    raw_text: Mapped[str] = mapped_column(Text, default="")
+    title: Mapped[str] = mapped_column(String(500), default="")
+    status: Mapped[str] = mapped_column(String(32), default="open")
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    closed_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+
+
+class Document(Base):
+    __tablename__ = "documents"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    user_id: Mapped[int] = mapped_column(ForeignKey("users.id", ondelete="CASCADE"), index=True)
+    title: Mapped[str] = mapped_column(String(255), default="")
+    filename: Mapped[str] = mapped_column(String(255), default="")
+    content_hash: Mapped[str] = mapped_column(String(64), default="", index=True)
+    size_bytes: Mapped[int] = mapped_column(Integer, default=0)
+    created_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now()
+    )
+
+    chunks: Mapped[list["DocumentChunk"]] = relationship(
+        back_populates="document",
+        cascade="all, delete-orphan",
+        order_by="DocumentChunk.chunk_index",
+    )
+
+
+class DocumentChunk(Base):
+    __tablename__ = "document_chunks"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    document_id: Mapped[int] = mapped_column(
+        ForeignKey("documents.id", ondelete="CASCADE"), index=True
+    )
+    chunk_index: Mapped[int] = mapped_column(Integer, default=0)
+    content: Mapped[str] = mapped_column(Text, default="")
+
+    document: Mapped["Document"] = relationship(back_populates="chunks")

backend/app/db/models.py.refactor_bak

@@ -0,0 +1,299 @@
+from datetime import datetime
+
+from sqlalchemy import Boolean, DateTime, Float, ForeignKey, Integer, String, Text, func
+from sqlalchemy.orm import Mapped, mapped_column, relationship
+
+from app.db.base import Base
+
+
+class ChatSession(Base):
+    __tablename__ = "chat_sessions"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    title: Mapped[str] = mapped_column(String(255), default="Новый чат")
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+    messages: Mapped[list["Message"]] = relationship(
+        back_populates="session", cascade="all, delete-orphan", order_by="Message.created_at"
+    )
+
+
+class Message(Base):
+    __tablename__ = "messages"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    session_id: Mapped[int] = mapped_column(ForeignKey("chat_sessions.id", ondelete="CASCADE"), index=True)
+    role: Mapped[str] = mapped_column(String(32))
+    content: Mapped[str] = mapped_column(Text, default="")
+    tool_calls_json: Mapped[str | None] = mapped_column(Text, nullable=True)
+    reasoning_json: Mapped[str | None] = mapped_column(Text, nullable=True)
+    tool_call_id: Mapped[str | None] = mapped_column(String(64), nullable=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+    session: Mapped["ChatSession"] = relationship(back_populates="messages")
+
+
+class PomodoroCycle(Base):
+    __tablename__ = "pomodoro_cycles"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    work_duration_min: Mapped[int] = mapped_column(Integer, default=25)
+    short_break_min: Mapped[int] = mapped_column(Integer, default=5)
+    long_break_min: Mapped[int] = mapped_column(Integer, default=15)
+    sessions_until_long_break: Mapped[int] = mapped_column(Integer, default=4)
+    completed_work_sessions: Mapped[int] = mapped_column(Integer, default=0)
+    task_note: Mapped[str] = mapped_column(Text, default="")
+    auto_advance: Mapped[bool] = mapped_column(Boolean, default=True)
+    chat_notify_seq: Mapped[int] = mapped_column(Integer, default=0)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class PomodoroSession(Base):
+    __tablename__ = "pomodoro_sessions"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    status: Mapped[str] = mapped_column(String(32), default="idle")
+    phase: Mapped[str] = mapped_column(String(32), default="work")
+    duration_min: Mapped[int] = mapped_column(Integer, default=25)
+    task_note: Mapped[str] = mapped_column(Text, default="")
+    result: Mapped[str | None] = mapped_column(Text, nullable=True)
+    completed: Mapped[bool] = mapped_column(Boolean, default=False)
+    completion_notified: Mapped[bool] = mapped_column(Boolean, default=False)
+    started_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    paused_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    elapsed_seconds: Mapped[int] = mapped_column(Integer, default=0)
+    finished_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+
+class TaigaProject(Base):
+    __tablename__ = "taiga_projects"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    taiga_id: Mapped[int] = mapped_column(Integer, unique=True, index=True)
+    name: Mapped[str] = mapped_column(String(255))
+    slug: Mapped[str] = mapped_column(String(255), unique=True, index=True)
+    synced_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+
+class ProjectBinding(Base):
+    __tablename__ = "project_bindings"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    taiga_slug: Mapped[str] = mapped_column(String(255), unique=True, index=True)
+    gitea_owner: Mapped[str] = mapped_column(String(255), default="")
+    gitea_repo: Mapped[str] = mapped_column(String(255), default="")
+    default_branch: Mapped[str] = mapped_column(String(64), default="main")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class UserProfile(Base):
+    __tablename__ = "user_profile"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    data_json: Mapped[str] = mapped_column(Text, default="{}")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class MemoryFact(Base):
+    __tablename__ = "memory_facts"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    category: Mapped[str] = mapped_column(String(64), default="fact", index=True)
+    content: Mapped[str] = mapped_column(Text)
+    source: Mapped[str] = mapped_column(String(32), default="user")
+    session_id: Mapped[int | None] = mapped_column(
+        ForeignKey("chat_sessions.id", ondelete="SET NULL"), nullable=True, index=True
+    )
+    importance: Mapped[int] = mapped_column(Integer, default=3)
+    active: Mapped[bool] = mapped_column(Boolean, default=True, index=True)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class SessionSummary(Base):
+    __tablename__ = "session_summaries"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    session_id: Mapped[int] = mapped_column(
+        ForeignKey("chat_sessions.id", ondelete="CASCADE"), unique=True, index=True
+    )
+    summary: Mapped[str] = mapped_column(Text, default="")
+    message_count: Mapped[int] = mapped_column(Integer, default=0)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class FitnessProfile(Base):
+    __tablename__ = "fitness_profiles"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    sex: Mapped[str] = mapped_column(String(16), default="male")
+    age: Mapped[int] = mapped_column(Integer, default=30)
+    height_cm: Mapped[float] = mapped_column(Float, default=170.0)
+    weight_kg: Mapped[float] = mapped_column(Float, default=70.0)
+    activity_level: Mapped[str] = mapped_column(String(32), default="moderate")
+    goal: Mapped[str] = mapped_column(String(32), default="maintain")
+    target_weight_kg: Mapped[float | None] = mapped_column(Float, nullable=True)
+    weekly_workouts: Mapped[int] = mapped_column(Integer, default=3)
+    calorie_target: Mapped[float] = mapped_column(Float, default=2000.0)
+    protein_g: Mapped[float] = mapped_column(Float, default=140.0)
+    fat_g: Mapped[float] = mapped_column(Float, default=65.0)
+    carbs_g: Mapped[float] = mapped_column(Float, default=200.0)
+    water_l: Mapped[float] = mapped_column(Float, default=2.5)
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class BodyMetric(Base):
+    __tablename__ = "body_metrics"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    recorded_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    weight_kg: Mapped[float] = mapped_column(Float)
+    body_fat_pct: Mapped[float | None] = mapped_column(Float, nullable=True)
+    chest_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
+    waist_cm: Mapped[float | None] = mapped_column(Float, nullable=True)
+    notes: Mapped[str] = mapped_column(Text, default="")
+
+
+class FoodLog(Base):
+    __tablename__ = "food_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    meal_type: Mapped[str] = mapped_column(String(32), default="snack")
+    description: Mapped[str] = mapped_column(Text, default="")
+    calories: Mapped[float] = mapped_column(Float, default=0)
+    protein_g: Mapped[float] = mapped_column(Float, default=0)
+    fat_g: Mapped[float] = mapped_column(Float, default=0)
+    carbs_g: Mapped[float] = mapped_column(Float, default=0)
+    source: Mapped[str] = mapped_column(String(32), default="llm")
+    estimated: Mapped[bool] = mapped_column(Boolean, default=True)
+
+
+class WaterLog(Base):
+    __tablename__ = "water_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    amount_ml: Mapped[int] = mapped_column(Integer)
+
+
+class WorkoutLog(Base):
+    __tablename__ = "workout_logs"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    logged_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    title: Mapped[str] = mapped_column(String(255), default="Тренировка")
+    notes: Mapped[str] = mapped_column(Text, default="")
+    duration_min: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    exercises_json: Mapped[str] = mapped_column(Text, default="[]")
+
+
+class FitnessReminder(Base):
+    __tablename__ = "fitness_reminders"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    kind: Mapped[str] = mapped_column(String(32))
+    hour: Mapped[int] = mapped_column(Integer, default=12)
+    minute: Mapped[int] = mapped_column(Integer, default=0)
+    interval_hours: Mapped[int | None] = mapped_column(Integer, nullable=True)
+    enabled: Mapped[bool] = mapped_column(Boolean, default=True)
+    last_fired_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+
+
+class ShoppingList(Base):
+    __tablename__ = "shopping_lists"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    name: Mapped[str] = mapped_column(String(255), unique=True, index=True)
+    sort_order: Mapped[int] = mapped_column(Integer, default=0)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+    items: Mapped[list["ShoppingListItem"]] = relationship(
+        back_populates="shopping_list",
+        cascade="all, delete-orphan",
+        order_by="ShoppingListItem.sort_order, ShoppingListItem.id",
+    )
+
+
+class ShoppingListItem(Base):
+    __tablename__ = "shopping_list_items"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    list_id: Mapped[int] = mapped_column(
+        ForeignKey("shopping_lists.id", ondelete="CASCADE"), index=True
+    )
+    text: Mapped[str] = mapped_column(String(500))
+    quantity: Mapped[float | None] = mapped_column(Float, nullable=True)
+    unit: Mapped[str] = mapped_column(String(64), default="")
+    checked: Mapped[bool] = mapped_column(Boolean, default=False, index=True)
+    sort_order: Mapped[int] = mapped_column(Integer, default=0)
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+
+    shopping_list: Mapped["ShoppingList"] = relationship(back_populates="items")
+
+
+class Reminder(Base):
+    __tablename__ = "reminders"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    title: Mapped[str] = mapped_column(String(255))
+    notes: Mapped[str] = mapped_column(Text, default="")
+    due_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), index=True)
+    all_day: Mapped[bool] = mapped_column(Boolean, default=False)
+    recurrence: Mapped[str] = mapped_column(String(16), default="none")
+    enabled: Mapped[bool] = mapped_column(Boolean, default=True, index=True)
+    last_fired_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    completed_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)
+    timezone: Mapped[str] = mapped_column(String(64), default="Europe/Moscow")
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class AssistantState(Base):
+    __tablename__ = "assistant_state"
+
+    key: Mapped[str] = mapped_column(String(128), primary_key=True)
+    value: Mapped[str] = mapped_column(Text, default="")
+    updated_at: Mapped[datetime] = mapped_column(
+        DateTime(timezone=True), server_default=func.now(), onupdate=func.now()
+    )
+
+
+class WorkItem(Base):
+    __tablename__ = "work_items"
+
+    id: Mapped[int] = mapped_column(Integer, primary_key=True, autoincrement=True)
+    taiga_slug: Mapped[str] = mapped_column(String(255), index=True)
+    taiga_project_id: Mapped[int] = mapped_column(Integer)
+    taiga_story_id: Mapped[int] = mapped_column(Integer)
+    taiga_story_ref: Mapped[int] = mapped_column(Integer, index=True)
+    gitea_owner: Mapped[str] = mapped_column(String(255), default="")
+    gitea_repo: Mapped[str] = mapped_column(String(255), default="")
+    gitea_issue_number: Mapped[int | None] = mapped_column(Integer, nullable=True, index=True)
+    suggested_branch: Mapped[str] = mapped_column(String(255), default="")
+    raw_text: Mapped[str] = mapped_column(Text, default="")
+    title: Mapped[str] = mapped_column(String(500), default="")
+    status: Mapped[str] = mapped_column(String(32), default="open")
+    created_at: Mapped[datetime] = mapped_column(DateTime(timezone=True), server_default=func.now())
+    closed_at: Mapped[datetime | None] = mapped_column(DateTime(timezone=True), nullable=True)

backend/app/fitness/activity_budget.py

@@ -0,0 +1,143 @@
+from __future__ import annotations
+
+from dataclasses import asdict, dataclass
+from typing import Any
+
+BASELINE_STEPS_BY_LEVEL: dict[str, int] = {
+    "sedentary": 5000,
+    "light": 7000,
+    "moderate": 9000,
+    "active": 11000,
+    "very_active": 13000,
+}
+
+WORKOUT_KCAL_PER_SESSION = 200
+KCAL_PER_STEP_PER_KG = 0.0005
+FALLBACK_KCAL_PER_MIN = 6
+
+
+@dataclass
+class ActivityBonus:
+    steps: int
+    steps_baseline: int
+    steps_bonus_kcal: float
+    workout_active_kcal: float
+    workout_baseline_kcal: float
+    workout_bonus_kcal: float
+    total_bonus_kcal: float
+    scale_factor: float
+
+    def to_dict(self) -> dict[str, Any]:
+        return asdict(self)
+
+
+def baseline_steps(profile: dict[str, Any]) -> int:
+    override = profile.get("baseline_steps")
+    if override is not None:
+        return int(override)
+    level = str(profile.get("activity_level") or "moderate")
+    return BASELINE_STEPS_BY_LEVEL.get(level, 9000)
+
+
+def baseline_workout_kcal_day(profile: dict[str, Any]) -> float:
+    override = profile.get("baseline_workout_kcal")
+    if override is not None:
+        return float(override)
+    weekly = int(profile.get("weekly_workouts") or 3)
+    return round(weekly * WORKOUT_KCAL_PER_SESSION / 7, 1)
+
+
+def estimate_workout_active_kcal(workout: dict[str, Any]) -> float:
+    active = workout.get("active_calories")
+    if active is not None:
+        return float(active)
+    duration = workout.get("duration_min")
+    if duration:
+        return float(duration) * FALLBACK_KCAL_PER_MIN
+    return 0.0
+
+
+def steps_bonus_kcal(*, steps: int, baseline_steps: int, weight_kg: float) -> float:
+    extra_steps = max(0, steps - baseline_steps)
+    return round(extra_steps * weight_kg * KCAL_PER_STEP_PER_KG, 1)
+
+
+def compute_activity_bonus(
+    profile: dict[str, Any],
+    *,
+    steps_total: int,
+    workouts: list[dict[str, Any]],
+) -> ActivityBonus:
+    weight_kg = float(profile.get("weight_kg") or 70)
+    steps_base = baseline_steps(profile)
+    workout_base = baseline_workout_kcal_day(profile)
+
+    s_bonus = steps_bonus_kcal(steps=steps_total, baseline_steps=steps_base, weight_kg=weight_kg)
+    workout_active = round(sum(estimate_workout_active_kcal(w) for w in workouts), 1)
+    w_bonus = max(0.0, round(workout_active - workout_base, 1))
+    total_bonus = round(s_bonus + w_bonus, 1)
+
+    base_cal = float(profile.get("calorie_target") or 2000)
+    scale_factor = 1.0 if base_cal <= 0 else round((base_cal + total_bonus) / base_cal, 4)
+
+    return ActivityBonus(
+        steps=steps_total,
+        steps_baseline=steps_base,
+        steps_bonus_kcal=s_bonus,
+        workout_active_kcal=workout_active,
+        workout_baseline_kcal=workout_base,
+        workout_bonus_kcal=w_bonus,
+        total_bonus_kcal=total_bonus,
+        scale_factor=scale_factor,
+    )
+
+
+def _targets_dict(
+    *,
+    calories: float,
+    protein_g: float,
+    fat_g: float,
+    carbs_g: float,
+    water_ml: float,
+) -> dict[str, float]:
+    return {
+        "calories": round(calories),
+        "protein_g": round(protein_g),
+        "fat_g": round(fat_g),
+        "carbs_g": round(carbs_g),
+        "water_ml": round(water_ml),
+    }
+
+
+def build_base_targets(profile: dict[str, Any]) -> dict[str, float]:
+    water_l = float(profile.get("water_l") or 2.5)
+    return _targets_dict(
+        calories=float(profile.get("calorie_target") or 2000),
+        protein_g=float(profile.get("protein_g") or 140),
+        fat_g=float(profile.get("fat_g") or 65),
+        carbs_g=float(profile.get("carbs_g") or 200),
+        water_ml=water_l * 1000,
+    )
+
+
+def scale_targets(
+    base_targets: dict[str, float],
+    bonus_kcal: float,
+) -> tuple[dict[str, float], dict[str, float]]:
+    """Return (effective_targets, targets_base). Water is not scaled."""
+    targets_base = dict(base_targets)
+    base_cal = float(base_targets["calories"])
+
+    if bonus_kcal <= 0 or base_cal <= 0:
+        return dict(base_targets), targets_base
+
+    scale = (base_cal + bonus_kcal) / base_cal
+    effective = _targets_dict(
+        calories=base_cal + bonus_kcal,
+        protein_g=float(base_targets["protein_g"]) * scale,
+        fat_g=float(base_targets["fat_g"]) * scale,
+        carbs_g=float(base_targets["carbs_g"]) * scale,
+        water_ml=float(base_targets["water_ml"]),
+    )
+    return effective, targets_base
+

backend/app/fitness/body_composition.py

@@ -0,0 +1,128 @@
+import math
+from typing import Any
+
+
+def _is_female(sex: str) -> bool:
+    return sex.lower() in ("f", "female", "ж", "женский", "woman")
+
+
+def _cm_to_inches(cm: float) -> float:
+    return cm / 2.54
+
+
+def _clamp_bf(value: float) -> float:
+    return round(max(3.0, min(50.0, value)), 1)
+
+
+def navy_body_fat_pct(
+    *,
+    sex: str,
+    height_cm: float,
+    neck_cm: float,
+    waist_cm: float,
+    hip_cm: float | None = None,
+) -> float | None:
+    if height_cm <= 0 or neck_cm <= 0 or waist_cm <= 0:
+        return None
+
+    height_in = _cm_to_inches(height_cm)
+    neck_in = _cm_to_inches(neck_cm)
+    waist_in = _cm_to_inches(waist_cm)
+
+    if _is_female(sex):
+        if hip_cm is None or hip_cm <= 0:
+            return None
+        hip_in = _cm_to_inches(hip_cm)
+        sum_in = waist_in + hip_in - neck_in
+        if sum_in <= 0:
+            return None
+        denom = (
+            1.29579
+            - 0.35004 * math.log10(sum_in)
+            + 0.22100 * math.log10(height_in)
+        )
+    else:
+        diff_in = waist_in - neck_in
+        if diff_in <= 0:
+            return None
+        denom = (
+            1.0324
+            - 0.19077 * math.log10(diff_in)
+            + 0.15456 * math.log10(height_in)
+        )
+
+    if denom <= 0:
+        return None
+
+    return _clamp_bf(495.0 / denom - 450.0)
+
+
+def whr(waist_cm: float, hip_cm: float) -> float | None:
+    if waist_cm <= 0 or hip_cm <= 0:
+        return None
+    return round(waist_cm / hip_cm, 2)
+
+
+def lean_body_mass(weight_kg: float, body_fat_pct: float) -> float:
+    return round(weight_kg * (1.0 - body_fat_pct / 100.0), 1)
+
+
+def ffmi(weight_kg: float, height_cm: float, body_fat_pct: float) -> float | None:
+    if height_cm <= 0:
+        return None
+    height_m = height_cm / 100.0
+    lbm = weight_kg * (1.0 - body_fat_pct / 100.0)
+    raw = lbm / (height_m * height_m)
+    normalized = raw + 6.1 * (1.8 - height_m)
+    return round(normalized, 1)
+
+
+def compute_body_composition(
+    *,
+    sex: str,
+    height_cm: float,
+    weight_kg: float,
+    neck_cm: float | None = None,
+    waist_cm: float | None = None,
+    hip_cm: float | None = None,
+    body_fat_pct: float | None = None,
+) -> dict[str, Any]:
+    warnings: list[str] = []
+    result: dict[str, Any] = {
+        "body_fat_pct": None,
+        "body_fat_method": None,
+        "whr": None,
+        "lbm_kg": None,
+        "ffmi": None,
+        "warnings": warnings,
+    }
+
+    bf = body_fat_pct
+    method: str | None = "manual" if bf is not None else None
+
+    if bf is None and neck_cm and waist_cm:
+        navy_bf = navy_body_fat_pct(
+            sex=sex,
+            height_cm=height_cm,
+            neck_cm=neck_cm,
+            waist_cm=waist_cm,
+            hip_cm=hip_cm,
+        )
+        if navy_bf is not None:
+            bf = navy_bf
+            method = "navy"
+        elif _is_female(sex) and not hip_cm:
+            warnings.append("Для Navy у женщин нужен обхват бёдер (hip_cm).")
+        elif neck_cm and waist_cm and waist_cm <= neck_cm:
+            warnings.append("Обхват талии должен быть больше шеи для Navy.")
+
+    if bf is not None:
+        result["body_fat_pct"] = round(float(bf), 1)
+        result["body_fat_method"] = method
+        result["lbm_kg"] = lean_body_mass(weight_kg, float(bf))
+        result["ffmi"] = ffmi(weight_kg, height_cm, float(bf))
+
+    if waist_cm and hip_cm:
+        result["whr"] = whr(waist_cm, hip_cm)
+
+    return result

backend/app/fitness/context.py

@@ -1,55 +1,94 @@
-from typing import Any
-
-from sqlalchemy.orm import Session
-
-from app.fitness.service import FitnessService
-
-
-def get_fitness_snapshot(db: Session) -> dict[str, Any]:
-    return FitnessService(db).snapshot()
-
-
-def format_fitness_context(snapshot: dict[str, Any]) -> str:
-    lines = ["[Фитнес — сводка на сегодня]"]
-
-    profile = snapshot.get("profile")
-    if not profile:
-        lines.append("Профиль не настроен. set_fitness_profile для целей ккал/БЖУ/воды.")
-    else:
-        lines.append(
-            f"Цели: {profile.get('calorie_target')} ккал, "
-            f"Б {profile.get('protein_g')} / Ж {profile.get('fat_g')} / У {profile.get('carbs_g')} г, "
-            f"вода {profile.get('water_l')} л"
-        )
-        if profile.get("goal"):
-            lines.append(
-                f"Цель: {profile.get('goal')}, вес {profile.get('weight_kg')} кг, "
-                f"рост {profile.get('height_cm')} см"
-            )
-
-    today = snapshot.get("today") or {}
-    totals = today.get("totals") or {}
-    targets = today.get("targets") or {}
-    water_l = totals.get("water_ml", 0) / 1000
-    water_target = targets.get("water_ml", 2500) / 1000
-
-    lines.append("")
-    lines.append(
-        f"Съедено: {totals.get('calories', 0):.0f}/{targets.get('calories', 0):.0f} ккал · "
-        f"Б {totals.get('protein_g', 0):.0f}/{targets.get('protein_g', 0):.0f} · "
-        f"Ж {totals.get('fat_g', 0):.0f}/{targets.get('fat_g', 0):.0f} · "
-        f"У {totals.get('carbs_g', 0):.0f}/{targets.get('carbs_g', 0):.0f} г"
-    )
-    lines.append(f"Вода: {water_l:.1f}/{water_target:.1f} л")
-
-    workouts = today.get("workouts") or []
-    if workouts:
-        lines.append(f"Тренировок сегодня: {len(workouts)}")
-
-    lines.append("")
-    lines.append(
-        "Правила: log_meal, log_water, log_weight, log_workout, get_fitness_summary (date/days_ago), get_fitness_history, "
-        "set_fitness_profile, calc_fitness_targets, lookup_food, lookup_exercise. "
-        "Еда — оценка LLM (≈), пользователь может уточнить."
-    )
-    return "\n".join(lines)
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.fitness.service import FitnessService
+
+
+def get_fitness_snapshot(db: Session, user_id: int) -> dict[str, Any]:
+    return FitnessService(db, user_id).snapshot()
+
+
+def format_fitness_context(snapshot: dict[str, Any]) -> str:
+    lines = ["[Фитнес — сводка на сегодня]"]
+
+    profile = snapshot.get("profile")
+    if not profile:
+        lines.append("Профиль не настроен. set_fitness_profile для целей ккал/БЖУ/воды.")
+    else:
+        lines.append(
+            f"Цели (база): {profile.get('calorie_target')} ккал, "
+            f"Б {profile.get('protein_g')} / Ж {profile.get('fat_g')} / У {profile.get('carbs_g')} г, "
+            f"вода {profile.get('water_l')} л"
+        )
+        if profile.get("goal"):
+            lines.append(
+                f"Цель: {profile.get('goal')}, вес {profile.get('weight_kg')} кг, "
+                f"рост {profile.get('height_cm')} см"
+            )
+
+    today = snapshot.get("today") or {}
+    totals = today.get("totals") or {}
+    targets = today.get("targets") or {}
+    targets_base = today.get("targets_base") or {}
+    activity = today.get("activity") or {}
+    steps_total = today.get("steps_total") or 0
+    water_l = totals.get("water_ml", 0) / 1000
+    water_target = targets.get("water_ml", 2500) / 1000
+
+    if profile and (activity.get("total_bonus_kcal") or steps_total):
+        lines.append(
+            f"Активность: шаги {steps_total} (база {activity.get('steps_baseline', 0)}), "
+            f"бонус +{activity.get('total_bonus_kcal', 0)} ккал"
+        )
+        base_cal = targets_base.get("calories", profile.get("calorie_target"))
+        lines.append(f"Эффективная цель ккал: {base_cal} → {targets.get('calories', base_cal)}")
+
+    lines.append("")
+    lines.append(
+        f"Съедено: {totals.get('calories', 0):.0f}/{targets.get('calories', 0):.0f} ккал · "
+        f"Б {totals.get('protein_g', 0):.0f}/{targets.get('protein_g', 0):.0f} · "
+        f"Ж {totals.get('fat_g', 0):.0f}/{targets.get('fat_g', 0):.0f} · "
+        f"У {totals.get('carbs_g', 0):.0f}/{targets.get('carbs_g', 0):.0f} г"
+    )
+    lines.append(f"Вода: {water_l:.1f}/{water_target:.1f} л")
+
+    workouts = today.get("workouts") or []
+    if workouts:
+        lines.append(f"Тренировок сегодня: {len(workouts)}")
+
+    stats = snapshot.get("workout_stats") or {}
+    if stats.get("count"):
+        lines.append(
+            f"Тренировки за {stats.get('days', 7)} дн.: {stats.get('count')} "
+            f"(цель/нед {stats.get('weekly_target')}, серия {stats.get('streak')} дн.)"
+        )
+
+    latest = (snapshot.get("body_metrics") or [None])[0]
+    if latest:
+        lines.append("")
+        lines.append("Антропометрия (последняя):")
+        parts = [f"{latest.get('weight_kg')} кг"]
+        if latest.get("body_fat_pct") is not None:
+            method = latest.get("body_fat_method") or "?"
+            parts.append(f"жир {latest.get('body_fat_pct')}% ({method})")
+        if latest.get("neck_cm"):
+            parts.append(f"шея {latest.get('neck_cm')}")
+        if latest.get("waist_cm"):
+            parts.append(f"талия {latest.get('waist_cm')}")
+        if latest.get("hip_cm"):
+            parts.append(f"бёдра {latest.get('hip_cm')}")
+        if latest.get("whr"):
+            parts.append(f"WHR {latest.get('whr')}")
+        if latest.get("ffmi"):
+            parts.append(f"FFMI {latest.get('ffmi')}")
+        lines.append(" · ".join(parts))
+
+    lines.append("")
+    lines.append(
+        "Правила: log_meal, log_water, log_weight (обхваты → Navy), log_steps, log_workout (date/days_ago), "
+        "calc_body_composition (расчёт без записи), get_fitness_summary (date/days_ago), get_fitness_history, "
+        "set_fitness_profile, calc_fitness_targets, lookup_food, lookup_exercise. "
+        "Еда — оценка LLM (≈), пользователь может уточнить."
+    )
+    return chr(10).join(lines)

backend/app/fitness/reminders.py

@@ -1,114 +1,111 @@
-from datetime import datetime, timedelta, timezone
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.config import get_settings
-from app.db.base import SessionLocal
-from app.db.models import ChatSession, FitnessReminder, Message
-from app.fitness.service import FitnessService
-
-KIND_LABELS = {
-    "water": "Вода",
-    "meal": "Еда",
-    "workout": "Тренировка",
-    "weigh_in": "Взвешивание",
-}
-
-
-def _post_fitness_notice(content: str) -> None:
-    db = SessionLocal()
-    try:
-        session = db.scalar(
-            select(ChatSession).order_by(ChatSession.updated_at.desc()).limit(1)
-        )
-        if not session:
-            session = ChatSession(title="Фитнес")
-            db.add(session)
-            db.commit()
-            db.refresh(session)
-        db.add(Message(session_id=session.id, role="notice", content=content))
-        db.commit()
-    finally:
-        db.close()
-
-
-def _build_notice(kind: str, summary: dict) -> str:
-    label = KIND_LABELS.get(kind, kind)
-    totals = summary.get("totals") or {}
-    targets = summary.get("targets") or {}
-    water_l = totals.get("water_ml", 0) / 1000
-    water_target = targets.get("water_ml", 2500) / 1000
-    cals = totals.get("calories", 0)
-    cal_target = targets.get("calories", 2000)
-
-    if kind == "water":
-        return (
-            f"💪 **{label}** · выпито {water_l:.1f}/{water_target:.1f} л сегодня. "
-            "Пора выпить стакан воды."
-        )
-    if kind == "meal":
-        return (
-            f"💪 **{label}** · {cals:.0f}/{cal_target:.0f} ккал за день. "
-            "Не забудь залогировать приём пищи."
-        )
-    if kind == "workout":
-        workouts = summary.get("workouts") or []
-        if workouts:
-            return f"💪 **{label}** · сегодня уже была тренировка. Отдыхай или лёгкая активность."
-        return "💪 **Тренировка** · запланирована на сегодня. Время двигаться!"
-    if kind == "weigh_in":
-        return "💪 **Взвешивание** · пора записать вес (log_weight)."
-    return f"💪 **{label}** · напоминание"
-
-
-def check_reminders(db: Session) -> list[str]:
-    if not get_settings().fitness_reminders_enabled:
-        return []
-
-    now = datetime.now(timezone.utc)
-    service = FitnessService(db)
-    summary = service.get_daily_summary()
-    fired: list[str] = []
-
-    reminders = db.scalars(
-        select(FitnessReminder).where(FitnessReminder.enabled.is_(True))
-    ).all()
-
-    for rem in reminders:
-        should_fire = False
-
-        if rem.interval_hours:
-            if rem.last_fired_at is None:
-                should_fire = now.hour >= rem.hour
-            else:
-                delta = now - rem.last_fired_at.replace(tzinfo=timezone.utc)
-                should_fire = delta >= timedelta(hours=rem.interval_hours)
-        else:
-            if rem.kind == "weigh_in":
-                if rem.last_fired_at:
-                    delta = now - rem.last_fired_at.replace(tzinfo=timezone.utc)
-                    should_fire = delta >= timedelta(days=7)
-                else:
-                    should_fire = now.hour == rem.hour and now.minute >= rem.minute
-            else:
-                if rem.last_fired_at:
-                    last = rem.last_fired_at.replace(tzinfo=timezone.utc)
-                    already_today = last.date() == now.date()
-                    if already_today:
-                        continue
-                should_fire = now.hour == rem.hour and now.minute >= rem.minute
-
-        if not should_fire:
-            continue
-
-        notice = _build_notice(rem.kind, summary)
-        rem.last_fired_at = now
-        fired.append(notice)
-
-    if fired:
-        db.commit()
-        for notice in fired:
-            _post_fitness_notice(notice)
-
-    return fired
+from datetime import datetime, timedelta, timezone
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.chat.notice_inbox import post_notice_to_latest_chat
+from app.config import get_settings
+from app.db.models import FitnessReminder, User
+from app.fitness.service import FitnessService
+
+KIND_LABELS = {
+    "water": "Вода",
+    "meal": "Еда",
+    "workout": "Тренировка",
+    "weigh_in": "Взвешивание",
+}
+
+
+def _build_notice(kind: str, summary: dict) -> str:
+    label = KIND_LABELS.get(kind, kind)
+    totals = summary.get("totals") or {}
+    targets = summary.get("targets") or {}
+    water_l = totals.get("water_ml", 0) / 1000
+    water_target = targets.get("water_ml", 2500) / 1000
+    cals = totals.get("calories", 0)
+    cal_target = targets.get("calories", 2000)
+
+    if kind == "water":
+        return (
+            f"💪 **{label}** · выпито {water_l:.1f}/{water_target:.1f} л сегодня. "
+            "Пора выпить стакан воды."
+        )
+    if kind == "meal":
+        return (
+            f"💪 **{label}** · {cals:.0f}/{cal_target:.0f} ккал за день. "
+            "Не забудь залогировать приём пищи."
+        )
+    if kind == "workout":
+        workouts = summary.get("workouts") or []
+        if workouts:
+            return f"💪 **{label}** · сегодня уже была тренировка. Отдыхай или лёгкая активность."
+        return "💪 **Тренировка** · запланирована на сегодня. Время двигаться!"
+    if kind == "weigh_in":
+        return "💪 **Взвешивание** · пора записать вес (log_weight)."
+    return f"💪 **{label}** · напоминание"
+
+
+def _check_user_reminders(db: Session, user_id: int) -> list[str]:
+    now = datetime.now(timezone.utc)
+    service = FitnessService(db, user_id)
+    summary = service.get_daily_summary()
+    fired: list[str] = []
+
+    reminders = db.scalars(
+        select(FitnessReminder).where(
+            FitnessReminder.user_id == user_id,
+            FitnessReminder.enabled.is_(True),
+        )
+    ).all()
+
+    for rem in reminders:
+        should_fire = False
+
+        if rem.interval_hours:
+            if rem.last_fired_at is None:
+                should_fire = now.hour >= rem.hour
+            else:
+                delta = now - rem.last_fired_at.replace(tzinfo=timezone.utc)
+                should_fire = delta >= timedelta(hours=rem.interval_hours)
+        else:
+            if rem.kind == "weigh_in":
+                if rem.last_fired_at:
+                    delta = now - rem.last_fired_at.replace(tzinfo=timezone.utc)
+                    should_fire = delta >= timedelta(days=7)
+                else:
+                    should_fire = now.hour == rem.hour and now.minute >= rem.minute
+            else:
+                if rem.last_fired_at:
+                    last = rem.last_fired_at.replace(tzinfo=timezone.utc)
+                    already_today = last.date() == now.date()
+                    if already_today:
+                        continue
+                should_fire = now.hour == rem.hour and now.minute >= rem.minute
+
+        if not should_fire:
+            continue
+
+        notice = _build_notice(rem.kind, summary)
+        rem.last_fired_at = now
+        fired.append(notice)
+
+    if fired:
+        for notice in fired:
+            post_notice_to_latest_chat(notice, user_id)
+
+    return fired
+
+
+def check_reminders(db: Session) -> list[str]:
+    if not get_settings().fitness_reminders_enabled:
+        return []
+
+    users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+    all_fired: list[str] = []
+    for user in users:
+        all_fired.extend(_check_user_reminders(db, user.id))
+
+    if all_fired:
+        db.commit()
+
+    return all_fired

backend/app/fitness/service.py.agent_read

@@ -0,0 +1,441 @@
+import json
+from datetime import date, datetime, time, timedelta, timezone
+from typing import Any
+
+from sqlalchemy import func, select
+from sqlalchemy.orm import Session
+
+from app.db.models import (
+    BodyMetric,
+    FitnessProfile,
+    FitnessReminder,
+    FoodLog,
+    WaterLog,
+    WorkoutLog,
+)
+from app.fitness.calculators import compute_targets, one_rep_max
+
+DEFAULT_REMINDERS = [
+    {"kind": "water", "hour": 9, "minute": 0, "interval_hours": 2},
+    {"kind": "meal", "hour": 13, "minute": 0, "interval_hours": None},
+    {"kind": "workout", "hour": 18, "minute": 0, "interval_hours": None},
+    {"kind": "weigh_in", "hour": 8, "minute": 0, "interval_hours": None},
+]
+
+
+class FitnessService:
+    def __init__(self, db: Session):
+        self.db = db
+
+    def _get_profile_row(self) -> FitnessProfile | None:
+        return self.db.scalar(select(FitnessProfile).limit(1))
+
+    def get_profile(self) -> dict[str, Any] | None:
+        row = self._get_profile_row()
+        if not row:
+            return None
+        return self._profile_to_dict(row)
+
+    def _profile_to_dict(self, row: FitnessProfile) -> dict[str, Any]:
+        targets = compute_targets(
+            {
+                "sex": row.sex,
+                "age": row.age,
+                "height_cm": row.height_cm,
+                "weight_kg": row.weight_kg,
+                "activity_level": row.activity_level,
+                "goal": row.goal,
+            }
+        )
+        return {
+            "sex": row.sex,
+            "age": row.age,
+            "height_cm": row.height_cm,
+            "weight_kg": row.weight_kg,
+            "activity_level": row.activity_level,
+            "goal": row.goal,
+            "target_weight_kg": row.target_weight_kg,
+            "weekly_workouts": row.weekly_workouts,
+            "calorie_target": row.calorie_target,
+            "protein_g": row.protein_g,
+            "fat_g": row.fat_g,
+            "carbs_g": row.carbs_g,
+            "water_l": row.water_l,
+            "computed": targets,
+            "updated_at": row.updated_at.isoformat() if row.updated_at else None,
+        }
+
+    def set_profile(self, updates: dict[str, Any]) -> dict[str, Any]:
+        row = self._get_profile_row()
+        is_new = row is None
+        if is_new:
+            row = FitnessProfile()
+            self.db.add(row)
+            self.db.flush()
+
+        for key in (
+            "sex", "age", "height_cm", "weight_kg", "activity_level",
+            "goal", "target_weight_kg", "weekly_workouts",
+        ):
+            if key in updates and updates[key] is not None:
+                setattr(row, key, updates[key])
+
+        targets = compute_targets(
+            {
+                "sex": row.sex,
+                "age": row.age,
+                "height_cm": row.height_cm,
+                "weight_kg": row.weight_kg,
+                "activity_level": row.activity_level,
+                "goal": row.goal,
+            }
+        )
+        row.calorie_target = targets["calorie_target"]
+        row.protein_g = targets["protein_g"]
+        row.fat_g = targets["fat_g"]
+        row.carbs_g = targets["carbs_g"]
+        row.water_l = targets["water_l"]
+        row.updated_at = datetime.now(timezone.utc)
+
+        if is_new:
+            self._ensure_default_reminders()
+
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "profile": self._profile_to_dict(row)}
+
+    def _ensure_default_reminders(self) -> None:
+        existing = self.db.scalars(select(FitnessReminder)).all()
+        if existing:
+            return
+        for item in DEFAULT_REMINDERS:
+            self.db.add(FitnessReminder(**item))
+
+    def calc_targets(self, params: dict[str, Any]) -> dict[str, Any]:
+        return compute_targets(params)
+
+    def _day_bounds(self, day: date | None = None) -> tuple[datetime, datetime]:
+        d = day or datetime.now(timezone.utc).date()
+        start = datetime.combine(d, time.min, tzinfo=timezone.utc)
+        end = datetime.combine(d, time.max, tzinfo=timezone.utc)
+        return start, end
+
+    def get_daily_summary(self, day: date | None = None) -> dict[str, Any]:
+        start, end = self._day_bounds(day)
+        profile = self.get_profile()
+
+        foods = self.db.scalars(
+            select(FoodLog)
+            .where(FoodLog.logged_at >= start, FoodLog.logged_at <= end)
+            .order_by(FoodLog.logged_at)
+        ).all()
+        waters = self.db.scalars(
+            select(WaterLog)
+            .where(WaterLog.logged_at >= start, WaterLog.logged_at <= end)
+            .order_by(WaterLog.logged_at)
+        ).all()
+        workouts = self.db.scalars(
+            select(WorkoutLog)
+            .where(WorkoutLog.logged_at >= start, WorkoutLog.logged_at <= end)
+            .order_by(WorkoutLog.logged_at)
+        ).all()
+
+        totals = {
+            "calories": sum(f.calories for f in foods),
+            "protein_g": sum(f.protein_g for f in foods),
+            "fat_g": sum(f.fat_g for f in foods),
+            "carbs_g": sum(f.carbs_g for f in foods),
+            "water_ml": sum(w.amount_ml for w in waters),
+        }
+
+        targets = profile or {
+            "calorie_target": 2000,
+            "protein_g": 140,
+            "fat_g": 65,
+            "carbs_g": 200,
+            "water_l": 2.5,
+        }
+
+        return {
+            "date": (day or datetime.now(timezone.utc).date()).isoformat(),
+            "profile_configured": profile is not None,
+            "totals": totals,
+            "targets": {
+                "calories": targets.get("calorie_target", 2000),
+                "protein_g": targets.get("protein_g", 140),
+                "fat_g": targets.get("fat_g", 65),
+                "carbs_g": targets.get("carbs_g", 200),
+                "water_ml": targets.get("water_l", 2.5) * 1000,
+            },
+            "meals": [self._food_to_dict(f) for f in foods],
+            "water": [self._water_to_dict(w) for w in waters],
+            "workouts": [self._workout_to_dict(w) for w in workouts],
+        }
+
+    def log_meal(
+        self,
+        *,
+        description: str,
+        meal_type: str = "snack",
+        calories: float = 0,
+        protein_g: float = 0,
+        fat_g: float = 0,
+        carbs_g: float = 0,
+        source: str = "llm",
+        estimated: bool = True,
+    ) -> dict[str, Any]:
+        row = FoodLog(
+            meal_type=meal_type[:32],
+            description=description[:2000],
+            calories=calories,
+            protein_g=protein_g,
+            fat_g=fat_g,
+            carbs_g=carbs_g,
+            source=source[:32],
+            estimated=estimated,
+        )
+        self.db.add(row)
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "meal": self._food_to_dict(row)}
+
+    def log_water(self, amount_ml: int) -> dict[str, Any]:
+        row = WaterLog(amount_ml=max(0, amount_ml))
+        self.db.add(row)
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "water": self._water_to_dict(row)}
+
+    def log_weight(
+        self,
+        weight_kg: float,
+        *,
+        body_fat_pct: float | None = None,
+        chest_cm: float | None = None,
+        waist_cm: float | None = None,
+        notes: str = "",
+    ) -> dict[str, Any]:
+        row = BodyMetric(
+            weight_kg=weight_kg,
+            body_fat_pct=body_fat_pct,
+            chest_cm=chest_cm,
+            waist_cm=waist_cm,
+            notes=notes[:1000],
+        )
+        self.db.add(row)
+        profile = self._get_profile_row()
+        if profile:
+            profile.weight_kg = weight_kg
+            targets = compute_targets(
+                {
+                    "sex": profile.sex,
+                    "age": profile.age,
+                    "height_cm": profile.height_cm,
+                    "weight_kg": weight_kg,
+                    "activity_level": profile.activity_level,
+                    "goal": profile.goal,
+                }
+            )
+            profile.calorie_target = targets["calorie_target"]
+            profile.protein_g = targets["protein_g"]
+            profile.fat_g = targets["fat_g"]
+            profile.carbs_g = targets["carbs_g"]
+            profile.water_l = targets["water_l"]
+        self.db.commit()
+        self.db.refresh(row)
+        return {
+            "ok": True,
+            "metric": {
+                "id": row.id,
+                "weight_kg": row.weight_kg,
+                "recorded_at": row.recorded_at.isoformat() if row.recorded_at else None,
+            },
+        }
+
+    def log_workout(
+        self,
+        *,
+        title: str,
+        notes: str = "",
+        duration_min: int | None = None,
+        exercises: list[dict[str, Any]] | None = None,
+    ) -> dict[str, Any]:
+        row = WorkoutLog(
+            title=title[:255],
+            notes=notes[:2000],
+            duration_min=duration_min,
+            exercises_json=json.dumps(exercises or [], ensure_ascii=False),
+        )
+        self.db.add(row)
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "workout": self._workout_to_dict(row)}
+
+    def list_body_metrics(self, limit: int = 30) -> list[dict[str, Any]]:
+        rows = self.db.scalars(
+            select(BodyMetric).order_by(BodyMetric.recorded_at.desc()).limit(limit)
+        ).all()
+        return [
+            {
+                "id": r.id,
+                "weight_kg": r.weight_kg,
+                "body_fat_pct": r.body_fat_pct,
+                "chest_cm": r.chest_cm,
+                "waist_cm": r.waist_cm,
+                "notes": r.notes,
+                "recorded_at": r.recorded_at.isoformat() if r.recorded_at else None,
+            }
+            for r in rows
+        ]
+
+    def delete_food_log(self, log_id: int) -> bool:
+        row = self.db.get(FoodLog, log_id)
+        if not row:
+            return False
+        self.db.delete(row)
+        self.db.commit()
+        return True
+
+    def delete_water_log(self, log_id: int) -> bool:
+        row = self.db.get(WaterLog, log_id)
+        if not row:
+            return False
+        self.db.delete(row)
+        self.db.commit()
+        return True
+
+    def delete_workout_log(self, log_id: int) -> bool:
+        row = self.db.get(WorkoutLog, log_id)
+        if not row:
+            return False
+        self.db.delete(row)
+        self.db.commit()
+        return True
+
+    def list_reminders(self) -> list[dict[str, Any]]:
+        rows = self.db.scalars(select(FitnessReminder).order_by(FitnessReminder.kind)).all()
+        return [self._reminder_to_dict(r) for r in rows]
+
+    def set_reminder(
+        self,
+        kind: str,
+        *,
+        enabled: bool | None = None,
+        hour: int | None = None,
+        minute: int | None = None,
+        interval_hours: int | None = None,
+    ) -> dict[str, Any]:
+        row = self.db.scalar(
+            select(FitnessReminder).where(FitnessReminder.kind == kind)
+        )
+        if not row:
+            row = FitnessReminder(kind=kind)
+            self.db.add(row)
+        if enabled is not None:
+            row.enabled = enabled
+        if hour is not None:
+            row.hour = hour
+        if minute is not None:
+            row.minute = minute
+        if interval_hours is not None:
+            row.interval_hours = interval_hours
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "reminder": self._reminder_to_dict(row)}
+
+    def calc_one_rm(self, weight_kg: float, reps: int) -> dict[str, Any]:
+        return {"ok": True, "one_rm_kg": one_rep_max(weight_kg, reps)}
+
+    def get_history(
+        self,
+        *,
+        days: int = 7,
+        end_day: date | None = None,
+    ) -> dict[str, Any]:
+        days = max(1, min(days, 90))
+        end = end_day or datetime.now(timezone.utc).date()
+        start = end - timedelta(days=days - 1)
+        summaries: list[dict[str, Any]] = []
+
+        for offset in range(days):
+            d = start + timedelta(days=offset)
+            full = self.get_daily_summary(d)
+            totals = full["totals"]
+            has_data = bool(full["meals"] or full["water"] or full["workouts"])
+            summaries.append(
+                {
+                    "date": full["date"],
+                    "has_data": has_data,
+                    "totals": totals,
+                    "targets": full["targets"],
+                    "meal_count": len(full["meals"]),
+                    "workout_count": len(full["workouts"]),
+                }
+            )
+
+        return {
+            "start_date": start.isoformat(),
+            "end_date": end.isoformat(),
+            "days": days,
+            "summaries": summaries,
+        }
+
+    def snapshot(self) -> dict[str, Any]:
+        today = datetime.now(timezone.utc).date()
+        return {
+            "profile": self.get_profile(),
+            "today": self.get_daily_summary(today),
+            "history": self.get_history(days=7, end_day=today),
+            "body_metrics": self.list_body_metrics(limit=10),
+            "reminders": self.list_reminders(),
+        }
+
+    @staticmethod
+    def _food_to_dict(row: FoodLog) -> dict[str, Any]:
+        return {
+            "id": row.id,
+            "meal_type": row.meal_type,
+            "description": row.description,
+            "calories": row.calories,
+            "protein_g": row.protein_g,
+            "fat_g": row.fat_g,
+            "carbs_g": row.carbs_g,
+            "source": row.source,
+            "estimated": row.estimated,
+            "logged_at": row.logged_at.isoformat() if row.logged_at else None,
+        }
+
+    @staticmethod
+    def _water_to_dict(row: WaterLog) -> dict[str, Any]:
+        return {
+            "id": row.id,
+            "amount_ml": row.amount_ml,
+            "logged_at": row.logged_at.isoformat() if row.logged_at else None,
+        }
+
+    @staticmethod
+    def _workout_to_dict(row: WorkoutLog) -> dict[str, Any]:
+        try:
+            exercises = json.loads(row.exercises_json or "[]")
+        except json.JSONDecodeError:
+            exercises = []
+        return {
+            "id": row.id,
+            "title": row.title,
+            "notes": row.notes,
+            "duration_min": row.duration_min,
+            "exercises": exercises,
+            "logged_at": row.logged_at.isoformat() if row.logged_at else None,
+        }
+
+    @staticmethod
+    def _reminder_to_dict(row: FitnessReminder) -> dict[str, Any]:
+        return {
+            "id": row.id,
+            "kind": row.kind,
+            "hour": row.hour,
+            "minute": row.minute,
+            "interval_hours": row.interval_hours,
+            "enabled": row.enabled,
+            "last_fired_at": row.last_fired_at.isoformat() if row.last_fired_at else None,
+        }

backend/app/fitness/structuring.py

@@ -1,66 +1,96 @@
-import json
-from typing import Any
-
-from app.llm.client import LLMClient
-from app.projects.structuring import strip_markdown_json
-
-MEAL_PROMPT = """
-Преобразуй описание еды в JSON. Только JSON, без markdown.
-Схема:
-{
-  "meal_type": "breakfast|lunch|dinner|snack",
-  "description": "краткое описание",
-  "calories": 0,
-  "protein_g": 0,
-  "fat_g": 0,
-  "carbs_g": 0,
-  "estimated": true
-}
-Правила:
-- Оцени ккал и БЖУ по типичным значениям для России/СНГ.
-- Все числа — float/int, метрическая система (г, ккал).
-- meal_type угадай из контекста или snack.
-- estimated всегда true для LLM-оценки.
-""".strip()
-
-WORKOUT_PROMPT = """
-Преобразуй описание тренировки в JSON. Только JSON.
-Схема:
-{
-  "title": "название",
-  "duration_min": null,
-  "notes": "",
-  "exercises": [
-    {"name": "жим лёжа", "sets": 3, "reps": 8, "weight_kg": 80}
-  ]
-}
-Правила:
-- weight_kg в кг, метрическая система.
-- Если данных нет — null или пустой массив.
-""".strip()
-
-
-async def structure_meal(raw_text: str) -> dict[str, Any]:
-    llm = LLMClient()
-    result = await llm.complete(
-        [
-            {"role": "system", "content": MEAL_PROMPT},
-            {"role": "user", "content": raw_text},
-        ],
-        temperature=0.2,
-    )
-    raw = strip_markdown_json(result.get("content") or "")
-    return json.loads(raw)
-
-
-async def structure_workout(raw_text: str) -> dict[str, Any]:
-    llm = LLMClient()
-    result = await llm.complete(
-        [
-            {"role": "system", "content": WORKOUT_PROMPT},
-            {"role": "user", "content": raw_text},
-        ],
-        temperature=0.2,
-    )
-    raw = strip_markdown_json(result.get("content") or "")
-    return json.loads(raw)
+import json
+from typing import Any
+
+from app.llm.client import LLMClient
+from app.projects.structuring import strip_markdown_json
+
+MEAL_PROMPT = """
+Преобразуй описание еды в JSON. Только JSON, без markdown.
+Схема:
+{
+  "meal_type": "breakfast|lunch|dinner|snack",
+  "description": "краткое описание",
+  "calories": 0,
+  "protein_g": 0,
+  "fat_g": 0,
+  "carbs_g": 0,
+  "estimated": true
+}
+Правила:
+- Оцени ккал и БЖУ по типичным значениям для России/СНГ.
+- Все числа — float/int, метрическая система (г, ккал).
+- meal_type угадай из контекста или snack.
+- estimated всегда true для LLM-оценки.
+""".strip()
+
+WORKOUT_PROMPT = """
+Преобразуй описание тренировки в JSON. Только JSON.
+Формат:
+{
+  "title": "название",
+  "duration_min": null,
+  "active_calories": null,
+  "total_calories": null,
+  "steps": null,
+  "notes": "",
+  "exercises": [
+    {"name": "имя упраж", "sets": 3, "reps": 8, "weight_kg": 80}
+  ]
+}
+Правила:
+- weight_kg в кг, округляй разумно.
+- active_calories / total_calories / steps — если упомянуты в тексте, иначе null.
+- Если данных нет — null или пустой массив.
+""".strip()
+
+STEPS_PROMPT = """
+Преобразуй запись о шагах в JSON. Только JSON.
+Формат:
+{
+  "steps": 0,
+  "active_calories": null,
+  "notes": ""
+}
+Правила:
+- steps — целое число шагов за день.
+- active_calories — только если явно указаны.
+""".strip()
+
+
+async def structure_meal(raw_text: str) -> dict[str, Any]:
+    llm = LLMClient()
+    result = await llm.complete(
+        [
+            {"role": "system", "content": MEAL_PROMPT},
+            {"role": "user", "content": raw_text},
+        ],
+        temperature=0.2,
+    )
+    raw = strip_markdown_json(result.get("content") or "")
+    return json.loads(raw)
+
+
+async def structure_workout(raw_text: str) -> dict[str, Any]:
+    llm = LLMClient()
+    result = await llm.complete(
+        [
+            {"role": "system", "content": WORKOUT_PROMPT},
+            {"role": "user", "content": raw_text},
+        ],
+        temperature=0.2,
+    )
+    raw = strip_markdown_json(result.get("content") or "")
+    return json.loads(raw)
+
+
+async def structure_steps(raw_text: str) -> dict[str, Any]:
+    llm = LLMClient()
+    result = await llm.complete(
+        [
+            {"role": "system", "content": STEPS_PROMPT},
+            {"role": "user", "content": raw_text},
+        ],
+        temperature=0.2,
+    )
+    raw = strip_markdown_json(result.get("content") or "")
+    return json.loads(raw)

backend/app/homelab/context.py

@@ -1,42 +1,42 @@
-from datetime import datetime
-from zoneinfo import ZoneInfo
-
-from sqlalchemy.orm import Session
-
-from app.memory.service import MemoryService
-
-WEEKDAY_RU = (
-    "понедельник",
-    "вторник",
-    "среда",
-    "четверг",
-    "пятница",
-    "суббота",
-    "воскресенье",
-)
-
-DEFAULT_TIMEZONE = "Europe/Moscow"
-
-
-def resolve_timezone(db: Session) -> str:
-    profile = MemoryService(db).get_profile()
-    tz = (profile.get("timezone") or "").strip()
-    return tz or DEFAULT_TIMEZONE
-
-
-def format_datetime_context(db: Session) -> str:
-    tz_name = resolve_timezone(db)
-    try:
-        tz = ZoneInfo(tz_name)
-    except Exception:
-        tz = ZoneInfo(DEFAULT_TIMEZONE)
-        tz_name = DEFAULT_TIMEZONE
-
-    now = datetime.now(tz)
-    weekday = WEEKDAY_RU[now.weekday()]
-    lines = [
-        "[Текущее время]",
-        f"Сейчас: {now.strftime('%Y-%m-%d %H:%M')} ({weekday}), часовой пояс {tz_name}.",
-        "Учитывай время при ответах о «сегодня», «утром», «вечером» и расписании.",
-    ]
-    return "\n".join(lines)
+from datetime import datetime
+from zoneinfo import ZoneInfo
+
+from sqlalchemy.orm import Session
+
+from app.memory.service import MemoryService
+
+WEEKDAY_RU = (
+    "понедельник",
+    "вторник",
+    "среда",
+    "четверг",
+    "пятница",
+    "суббота",
+    "воскресенье",
+)
+
+DEFAULT_TIMEZONE = "Europe/Moscow"
+
+
+def resolve_timezone(db: Session, user_id: int) -> str:
+    profile = MemoryService(db, user_id).get_profile()
+    tz = (profile.get("timezone") or "").strip()
+    return tz or DEFAULT_TIMEZONE
+
+
+def format_datetime_context(db: Session, user_id: int) -> str:
+    tz_name = resolve_timezone(db, user_id)
+    try:
+        tz = ZoneInfo(tz_name)
+    except Exception:
+        tz = ZoneInfo(DEFAULT_TIMEZONE)
+        tz_name = DEFAULT_TIMEZONE
+
+    now = datetime.now(tz)
+    weekday = WEEKDAY_RU[now.weekday()]
+    lines = [
+        "[Текущее время]",
+        f"Сейчас: {now.strftime('%Y-%m-%d %H:%M')} ({weekday}), часовой пояс {tz_name}.",
+        "Учитывай время при ответах о «сегодня», «утром», «вечером» и расписании.",
+    ]
+    return "\n".join(lines)

backend/app/homelab/image_gen.py

@@ -1,130 +1,131 @@
-from typing import Any
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.character.service import CharacterService
-from app.config import get_settings
-from app.db.models import Message
-from app.homelab.comfyui import ComfyUIClient
-from app.integrations.rp_chat import RpChatClient
-
-
-def _card_image_settings() -> dict[str, Any]:
-    return CharacterService().get_card().get("data", {})
-
-
-def _session_messages(db: Session, session_id: int | None, limit: int = 8) -> list[dict[str, str]]:
-    if not session_id:
-        return []
-    rows = db.scalars(
-        select(Message)
-        .where(
-            Message.session_id == session_id,
-            Message.role.in_(("user", "assistant")),
-        )
-        .order_by(Message.created_at.desc())
-        .limit(limit)
-    ).all()
-    rows = list(reversed(rows))
-    return [{"role": m.role, "content": (m.content or "").strip()} for m in rows if m.content.strip()]
-
-
-def _append_lora(positive: str, lora_name: str, lora_weight: float) -> str:
-    if not lora_name or f"<lora:{lora_name}" in positive:
-        return positive
-    return f"{positive} <lora:{lora_name}:{lora_weight}>"
-
-
-async def generate_image(
-    db: Session,
-    *,
-    session_id: int | None = None,
-    draw_self: bool = False,
-    scene_description: str = "",
-) -> dict[str, Any]:
-    card = _card_image_settings()
-    settings = get_settings()
-
-    if not card.get("sd_enabled", True):
-        return {"ok": False, "error": "Генерация изображений отключена в настройках персонажа"}
-
-    if not draw_self and not scene_description.strip():
-        return {"ok": False, "error": "Нужен draw_self=true или scene_description"}
-
-    appearance = (card.get("appearance_tags") or "").strip()
-    if draw_self and not appearance:
-        return {
-            "ok": False,
-            "error": "Заполни appearance_tags в настройках персонажа для «нарисуй себя»",
-        }
-
-    messages = _session_messages(db, session_id)
-    if scene_description.strip():
-        messages = messages + [{"role": "user", "content": scene_description.strip()}]
-    elif draw_self and messages:
-        messages = messages + [{"role": "user", "content": "Illustrate the current scene with the character."}]
-    elif draw_self:
-        messages = [{"role": "user", "content": "Portrait of the character, looking at viewer, friendly expression."}]
-
-    if settings.rp_chat_enabled:
-        appearance_override = (card.get("appearance_tags") or "").strip() or None
-        return await _generate_via_rp_chat(card, messages, appearance_override)
-
-    return await _generate_via_local_comfy(scene_description or "anime character portrait")
-
-
-async def _generate_via_rp_chat(
-    card: dict[str, Any],
-    messages: list[dict[str, str]],
-    appearance_override: str | None,
-) -> dict[str, Any]:
-    client = RpChatClient()
-    persona_id = (card.get("rp_persona_id") or "").strip() or "default"
-    override = appearance_override or (card.get("appearance_tags") or "").strip() or None
-
-    prompt_result = await client.sd_prompt(
-        persona_id,
-        messages,
-        appearance_override=override,
-    )
-    if not prompt_result.get("ok"):
-        return prompt_result
-
-    positive = (
-        prompt_result.get("hybrid_positive")
-        or prompt_result.get("tag_positive")
-        or ""
-    ).strip()
-    negative = (prompt_result.get("negative") or "").strip()
-    if not positive:
-        return {"ok": False, "error": "RP-чат не вернул промпт", "raw": prompt_result}
-
-    lora = (card.get("lora_name") or "").strip()
-    if lora:
-        weight = float(card.get("lora_weight") or 0.8)
-        positive = _append_lora(positive, lora, weight)
-
-    gen_result = await client.generate(positive, negative)
-    if not gen_result.get("ok"):
-        return gen_result
-
-    saved = await client.save_image_locally(gen_result["image_path"])
-    if not saved.get("ok"):
-        return saved
-
-    return {
-        "ok": True,
-        "url": saved["url"],
-        "filename": saved["filename"],
-        "prompt": positive,
-        "backend": "rp_chat",
-        "persona_id": persona_id,
-    }
-
-
-async def _generate_via_local_comfy(prompt: str) -> dict[str, Any]:
-    result = await ComfyUIClient().generate_image(prompt)
-    if result.get("ok"):
-        result["backend"] = "comfyui_local"
-    return result
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.character.service import CharacterService
+from app.config import get_settings
+from app.db.models import Message
+from app.homelab.comfyui import ComfyUIClient
+from app.integrations.rp_chat import RpChatClient
+
+
+def _card_image_settings(db: Session, user_id: int) -> dict[str, Any]:
+    return CharacterService(db, user_id).get_card().get("data", {})
+
+
+def _session_messages(db: Session, session_id: int | None, limit: int = 8) -> list[dict[str, str]]:
+    if not session_id:
+        return []
+    rows = db.scalars(
+        select(Message)
+        .where(
+            Message.session_id == session_id,
+            Message.role.in_(("user", "assistant")),
+        )
+        .order_by(Message.created_at.desc())
+        .limit(limit)
+    ).all()
+    rows = list(reversed(rows))
+    return [{"role": m.role, "content": (m.content or "").strip()} for m in rows if m.content.strip()]
+
+
+def _append_lora(positive: str, lora_name: str, lora_weight: float) -> str:
+    if not lora_name or f"<lora:{lora_name}" in positive:
+        return positive
+    return f"{positive} <lora:{lora_name}:{lora_weight}>"
+
+
+async def generate_image(
+    db: Session,
+    *,
+    user_id: int,
+    session_id: int | None = None,
+    draw_self: bool = False,
+    scene_description: str = "",
+) -> dict[str, Any]:
+    card = _card_image_settings(db, user_id)
+    settings = get_settings()
+
+    if not card.get("sd_enabled", True):
+        return {"ok": False, "error": "Генерация изображений отключена в настройках персонажа"}
+
+    if not draw_self and not scene_description.strip():
+        return {"ok": False, "error": "Нужен draw_self=true или scene_description"}
+
+    appearance = (card.get("appearance_tags") or "").strip()
+    if draw_self and not appearance:
+        return {
+            "ok": False,
+            "error": "Заполни appearance_tags в настройках персонажа для «нарисуй себя»",
+        }
+
+    messages = _session_messages(db, session_id)
+    if scene_description.strip():
+        messages = messages + [{"role": "user", "content": scene_description.strip()}]
+    elif draw_self and messages:
+        messages = messages + [{"role": "user", "content": "Illustrate the current scene with the character."}]
+    elif draw_self:
+        messages = [{"role": "user", "content": "Portrait of the character, looking at viewer, friendly expression."}]
+
+    if settings.rp_chat_enabled:
+        appearance_override = (card.get("appearance_tags") or "").strip() or None
+        return await _generate_via_rp_chat(card, messages, appearance_override)
+
+    return await _generate_via_local_comfy(scene_description or "anime character portrait")
+
+
+async def _generate_via_rp_chat(
+    card: dict[str, Any],
+    messages: list[dict[str, str]],
+    appearance_override: str | None,
+) -> dict[str, Any]:
+    client = RpChatClient()
+    persona_id = (card.get("rp_persona_id") or "").strip() or "default"
+    override = appearance_override or (card.get("appearance_tags") or "").strip() or None
+
+    prompt_result = await client.sd_prompt(
+        persona_id,
+        messages,
+        appearance_override=override,
+    )
+    if not prompt_result.get("ok"):
+        return prompt_result
+
+    positive = (
+        prompt_result.get("hybrid_positive")
+        or prompt_result.get("tag_positive")
+        or ""
+    ).strip()
+    negative = (prompt_result.get("negative") or "").strip()
+    if not positive:
+        return {"ok": False, "error": "RP-чат не вернул промпт", "raw": prompt_result}
+
+    lora = (card.get("lora_name") or "").strip()
+    if lora:
+        weight = float(card.get("lora_weight") or 0.8)
+        positive = _append_lora(positive, lora, weight)
+
+    gen_result = await client.generate(positive, negative)
+    if not gen_result.get("ok"):
+        return gen_result
+
+    saved = await client.save_image_locally(gen_result["image_path"])
+    if not saved.get("ok"):
+        return saved
+
+    return {
+        "ok": True,
+        "url": saved["url"],
+        "filename": saved["filename"],
+        "prompt": positive,
+        "backend": "rp_chat",
+        "persona_id": persona_id,
+    }
+
+
+async def _generate_via_local_comfy(prompt: str) -> dict[str, Any]:
+    result = await ComfyUIClient().generate_image(prompt)
+    if result.get("ok"):
+        result["backend"] = "comfyui_local"
+    return result

backend/app/homelab_scoped/notices.py

@@ -0,0 +1,5 @@
+from app.chat.notice_inbox import post_notice_to_latest_chat
+
+
+def post_chat_notice(content: str, user_id: int) -> None:
+    post_notice_to_latest_chat(content, user_id)

backend/app/homelab_scoped/watcher.py

@@ -0,0 +1,164 @@
+import asyncio
+import logging
+import random
+from datetime import datetime
+from zoneinfo import ZoneInfo
+
+import httpx
+from sqlalchemy import select
+
+from app.config import get_settings
+from app.db.base import SessionLocal
+from app.db.models import User
+from app.homelab.comfyui import ComfyUIClient
+from app.homelab.context import resolve_timezone
+from app.homelab.digest import build_morning_digest
+from app.homelab.monitoring import check_netdata_alerts
+from app.homelab_scoped.notices import post_chat_notice
+from app.homelab.state import get_state, set_state
+
+logger = logging.getLogger(__name__)
+
+WATCH_INTERVAL_SEC = 60
+_netdata_tick = 0
+
+
+async def homelab_watcher_loop() -> None:
+    global _netdata_tick
+    while True:
+        try:
+            await asyncio.sleep(WATCH_INTERVAL_SEC)
+            await _tick_morning_digest()
+            await _tick_rofl()
+            settings = get_settings()
+            _netdata_tick += WATCH_INTERVAL_SEC
+            if _netdata_tick >= settings.netdata_poll_interval_sec:
+                _netdata_tick = 0
+                await _tick_netdata()
+        except asyncio.CancelledError:
+            raise
+        except Exception:
+            logger.exception("Homelab watcher error")
+
+
+async def _tick_morning_digest() -> None:
+    settings = get_settings()
+    if not settings.morning_digest_enabled:
+        return
+
+    db = SessionLocal()
+    try:
+        users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+        digest = build_morning_digest(db, include_news=True)
+        for user in users:
+            tz_name = resolve_timezone(db, user.id)
+            try:
+                tz = ZoneInfo(tz_name)
+            except Exception:
+                tz = ZoneInfo("Europe/Moscow")
+
+            now = datetime.now(tz)
+            target_min = settings.morning_digest_hour * 60 + settings.morning_digest_minute
+            current_min = now.hour * 60 + now.minute
+            if current_min < target_min or current_min >= target_min + 3:
+                continue
+
+            today = now.date().isoformat()
+            state_key = f"last_morning_digest_date:{user.id}"
+            if get_state(db, state_key) == today:
+                continue
+
+            post_chat_notice(digest, user.id)
+            set_state(db, state_key, today)
+    finally:
+        db.close()
+
+
+async def _tick_netdata() -> None:
+    db = SessionLocal()
+    try:
+        notices = check_netdata_alerts(db)
+        if not notices:
+            return
+        users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+        for user in users:
+            for notice in notices:
+                post_chat_notice(notice, user.id)
+    finally:
+        db.close()
+
+
+async def _comfyui_reachable(base_url: str) -> bool:
+    try:
+        async with httpx.AsyncClient(timeout=httpx.Timeout(3.0, connect=2.0)) as client:
+            response = await client.get(f"{base_url.rstrip('/')}/system_stats")
+            return response.status_code < 500
+    except (httpx.TimeoutException, httpx.ConnectError, httpx.NetworkError):
+        return False
+
+
+async def _tick_rofl() -> None:
+    settings = get_settings()
+    if not settings.comfyui_enabled or not settings.comfyui_rofl_enabled:
+        return
+
+    db = SessionLocal()
+    try:
+        users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+        for user in users:
+            tz_name = resolve_timezone(db, user.id)
+            try:
+                tz = ZoneInfo(tz_name)
+            except Exception:
+                tz = ZoneInfo("Europe/Moscow")
+            now = datetime.now(tz)
+            last_raw = get_state(db, f"last_comfy_rofl_at:{user.id}")
+            if last_raw:
+                try:
+                    last_at = datetime.fromisoformat(last_raw)
+                    if last_at.tzinfo is None:
+                        last_at = last_at.replace(tzinfo=tz)
+                    if (now - last_at).total_seconds() < settings.comfyui_rofl_min_interval_hours * 3600:
+                        continue
+                except ValueError:
+                    pass
+
+            if random.random() > settings.comfyui_rofl_probability:
+                continue
+
+            today = now.date().isoformat()
+            count_key = f"comfy_rofl_count_{today}:{user.id}"
+            count_raw = get_state(db, count_key) or "0"
+            try:
+                count = int(count_raw)
+            except ValueError:
+                count = 0
+            if count >= settings.comfyui_rofl_max_per_day:
+                continue
+
+            client = ComfyUIClient()
+            if not await _comfyui_reachable(client.base_url):
+                continue
+
+            prompt = client.random_rofl_prompt()
+            try:
+                result = await asyncio.wait_for(
+                    client.generate_image(prompt),
+                    timeout=settings.comfyui_timeout_sec + 15,
+                )
+            except (asyncio.TimeoutError, httpx.TimeoutException, httpx.ConnectError) as exc:
+                logger.warning("Rofl image skipped (ComfyUI): %s", exc)
+                continue
+            if not result.get("ok"):
+                logger.warning("Rofl image failed: %s", result.get("error"))
+                continue
+
+            url = result.get("url", "")
+            post_chat_notice(
+                f"🎨 **Рофл дня**\n\n![rofl]({url})\n\n_{prompt}_",
+                user.id,
+            )
+            set_state(db, count_key, str(count + 1))
+            set_state(db, f"last_comfy_rofl_at:{user.id}", now.isoformat())
+    finally:
+        db.close()

backend/app/llm/client.py

@@ -1,269 +1,323 @@
-import json
-import logging
-from collections.abc import AsyncIterator
-from typing import Any
-
-from openai import AsyncOpenAI
-
-from app.config import get_settings
-
-logger = logging.getLogger(__name__)
-
-
-class LLMClient:
-    def __init__(self) -> None:
-        settings = get_settings()
-        self.model = settings.openrouter_model
-        self.tools_enabled = settings.openrouter_tools_enabled
-        self.reasoning_effort = settings.openrouter_reasoning_effort.strip().lower()
-        self.client = AsyncOpenAI(
-            api_key=settings.openrouter_api_key,
-            base_url=settings.openrouter_base_url,
-        )
-
-    def _reasoning_extra_body(self) -> dict[str, Any] | None:
-        if not self.reasoning_effort:
-            return None
-        if self.reasoning_effort == "none":
-            return {"reasoning": {"effort": "none", "exclude": True}}
-        return {"reasoning": {"effort": self.reasoning_effort}}
-
-    @staticmethod
-    def _delta_reasoning(delta: Any) -> tuple[str, list[Any]]:
-        parts: list[str] = []
-        for attr in ("reasoning", "reasoning_content"):
-            value = getattr(delta, attr, None)
-            if value:
-                parts.append(str(value))
-
-        details: list[Any] = []
-        raw_details = getattr(delta, "reasoning_details", None)
-        if raw_details:
-            if isinstance(raw_details, list):
-                details.extend(raw_details)
-            else:
-                details.append(raw_details)
-
-        return "".join(parts), details
-
-    @staticmethod
-    def _normalize_reasoning_details(details: Any) -> list[Any] | None:
-        if not details:
-            return None
-        items = details if isinstance(details, list) else [details]
-        normalized: list[Any] = []
-        for item in items:
-            if hasattr(item, "model_dump"):
-                normalized.append(item.model_dump())
-            elif isinstance(item, dict):
-                normalized.append(item)
-            else:
-                normalized.append(item)
-        return normalized or None
-
-    @staticmethod
-    def attach_reasoning_to_message(
-        message: dict[str, Any],
-        *,
-        reasoning: str = "",
-        reasoning_details: list[Any] | None = None,
-    ) -> dict[str, Any]:
-        if reasoning:
-            message["reasoning"] = reasoning
-            message["reasoning_content"] = reasoning
-        normalized = LLMClient._normalize_reasoning_details(reasoning_details)
-        if normalized:
-            message["reasoning_details"] = normalized
-        return message
-
-    async def stream_chat(
-        self,
-        messages: list[dict[str, Any]],
-        tools: list[dict[str, Any]] | None = None,
-        *,
-        model: str | None = None,
-    ) -> AsyncIterator[dict[str, Any]]:
-        use_tools = bool(tools) and self.tools_enabled
-        kwargs: dict[str, Any] = {
-            "model": model or self.model,
-            "messages": messages,
-            "stream": True,
-            "temperature": 0.7,
-        }
-        if use_tools:
-            kwargs["tools"] = tools
-        extra_body = self._reasoning_extra_body()
-        if extra_body:
-            kwargs["extra_body"] = extra_body
-
-        try:
-            stream = await self.client.chat.completions.create(**kwargs)
-        except Exception as exc:
-            logger.exception("LLM stream failed: %s", exc)
-            yield {"type": "error", "content": str(exc)}
-            yield {"type": "done", "finish_reason": "error"}
-            return
-
-        tool_calls: dict[int, dict[str, Any]] = {}
-        reasoning_parts: list[str] = []
-        reasoning_details: list[Any] = []
-
-        try:
-            async for chunk in stream:
-                if not chunk.choices:
-                    continue
-
-                choice = chunk.choices[0]
-                delta = choice.delta
-
-                if delta.content:
-                    yield {"type": "content", "content": delta.content}
-
-                reasoning_text, details = self._delta_reasoning(delta)
-                if reasoning_text:
-                    reasoning_parts.append(reasoning_text)
-                if details:
-                    reasoning_details.extend(details)
-
-                if delta.tool_calls:
-                    for tool_call in delta.tool_calls:
-                        idx = tool_call.index
-                        if idx not in tool_calls:
-                            tool_calls[idx] = {
-                                "id": tool_call.id or "",
-                                "type": "function",
-                                "function": {"name": "", "arguments": ""},
-                            }
-                        if tool_call.id:
-                            tool_calls[idx]["id"] = tool_call.id
-                        if tool_call.function:
-                            if tool_call.function.name:
-                                tool_calls[idx]["function"]["name"] = tool_call.function.name
-                            if tool_call.function.arguments:
-                                tool_calls[idx]["function"]["arguments"] += tool_call.function.arguments
-
-                if choice.finish_reason:
-                    reasoning = "".join(reasoning_parts)
-                    normalized_details = self._normalize_reasoning_details(reasoning_details)
-                    if reasoning or normalized_details:
-                        yield {
-                            "type": "reasoning",
-                            "reasoning": reasoning,
-                            "reasoning_details": normalized_details,
-                        }
-                    if tool_calls:
-                        yield {"type": "tool_calls", "tool_calls": list(tool_calls.values())}
-                    logger.info(
-                        "LLM stream done: model=%s finish_reason=%s tool_calls=%d "
-                        "content_in_stream=%d reasoning_len=%d",
-                        model or self.model,
-                        choice.finish_reason,
-                        len(tool_calls),
-                        len(reasoning_parts),
-                        len(reasoning),
-                    )
-                    yield {"type": "done", "finish_reason": choice.finish_reason}
-        except Exception as exc:
-            logger.exception("LLM stream read failed: %s", exc)
-            yield {"type": "error", "content": str(exc)}
-            yield {"type": "done", "finish_reason": "error"}
-
-    async def complete(
-        self,
-        messages: list[dict[str, Any]],
-        tools: list[dict[str, Any]] | None = None,
-        *,
-        temperature: float = 0.7,
-        model: str | None = None,
-        for_extraction: bool = False,
-        visible_reply: bool = False,
-    ) -> dict[str, Any]:
-        use_tools = bool(tools) and self.tools_enabled and not for_extraction
-        kwargs: dict[str, Any] = {
-            "model": model or self.model,
-            "messages": messages,
-            "temperature": temperature,
-        }
-        if use_tools:
-            kwargs["tools"] = tools
-        if for_extraction:
-            kwargs["extra_body"] = {"reasoning": {"effort": "none"}}
-        else:
-            extra_body = self._reasoning_extra_body()
-            if extra_body:
-                kwargs["extra_body"] = extra_body
-
-        response = await self.client.chat.completions.create(**kwargs)
-        message = response.choices[0].message
-
-        content = message.content or ""
-        reasoning = ""
-        for attr in ("reasoning", "reasoning_content"):
-            value = getattr(message, attr, None)
-            if value:
-                reasoning = str(value)
-                break
-
-        if not content and reasoning and not visible_reply:
-            content = reasoning
-
-        result: dict[str, Any] = {
-            "content": content,
-            "tool_calls": [],
-            "reasoning": reasoning,
-            "reasoning_details": getattr(message, "reasoning_details", None),
-        }
-
-        if message.tool_calls:
-            result["tool_calls"] = [
-                {
-                    "id": tc.id,
-                    "type": "function",
-                    "function": {
-                        "name": tc.function.name,
-                        "arguments": tc.function.arguments,
-                    },
-                }
-                for tc in message.tool_calls
-            ]
-
-        return result
-
-    @staticmethod
-    def parse_tool_arguments(arguments: str) -> dict[str, Any]:
-        if not arguments:
-            return {}
-        try:
-            return json.loads(arguments)
-        except json.JSONDecodeError:
-            return {}
-
-    @staticmethod
-    def serialize_reasoning(
-        *,
-        reasoning: str = "",
-        reasoning_details: list[Any] | None = None,
-    ) -> str | None:
-        payload: dict[str, Any] = {}
-        if reasoning:
-            payload["reasoning"] = reasoning
-            payload["reasoning_content"] = reasoning
-        if reasoning_details:
-            payload["reasoning_details"] = reasoning_details
-        if not payload:
-            return None
-        return json.dumps(payload, ensure_ascii=False)
-
-    @staticmethod
-    def deserialize_reasoning(raw: str | None) -> dict[str, Any]:
-        if not raw:
-            return {}
-        try:
-            data = json.loads(raw)
-        except json.JSONDecodeError:
-            return {"reasoning": raw}
-        if isinstance(data, str):
-            return {"reasoning": data, "reasoning_content": data}
-        if isinstance(data, dict):
-            return data
-        return {}
+import json
+import logging
+from collections.abc import AsyncIterator
+from typing import Any
+
+from openai import AsyncOpenAI
+
+from app.config import get_settings
+
+logger = logging.getLogger(__name__)
+
+
+class LLMClient:
+    def __init__(self) -> None:
+        settings = get_settings()
+        self.tools_enabled = settings.openrouter_tools_enabled
+        self.client = AsyncOpenAI(
+            api_key=settings.openrouter_api_key,
+            base_url=settings.openrouter_base_url,
+        )
+
+    def _runtime(self) -> tuple[str, str, str]:
+        from app.db.base import SessionLocal
+        from app.settings.service import SettingsService
+
+        settings = get_settings()
+        db = SessionLocal()
+        try:
+            svc = SettingsService(db)
+            model = str(svc.get_effective("openrouter_model"))
+            extract = str(svc.get_effective("memory_extract_model"))
+            effort = str(svc.get_effective("openrouter_reasoning_effort")).strip().lower()
+            return model, extract, effort
+        finally:
+            db.close()
+
+    @property
+    def model(self) -> str:
+        return self._runtime()[0]
+
+    @property
+    def memory_extract_model(self) -> str:
+        return self._runtime()[1]
+
+    @property
+    def reasoning_effort(self) -> str:
+        return self._runtime()[2]
+
+    def _reasoning_extra_body(self) -> dict[str, Any] | None:
+        if not self.reasoning_effort:
+            return None
+        if self.reasoning_effort == "none":
+            return {"reasoning": {"effort": "none", "exclude": True}}
+        return {"reasoning": {"effort": self.reasoning_effort}}
+
+    @staticmethod
+    def _delta_reasoning(delta: Any) -> tuple[str, list[Any]]:
+        parts: list[str] = []
+        for attr in ("reasoning", "reasoning_content"):
+            value = getattr(delta, attr, None)
+            if value:
+                parts.append(str(value))
+
+        details: list[Any] = []
+        raw_details = getattr(delta, "reasoning_details", None)
+        if raw_details:
+            if isinstance(raw_details, list):
+                details.extend(raw_details)
+            else:
+                details.append(raw_details)
+
+        return "".join(parts), details
+
+    @staticmethod
+    def _normalize_reasoning_details(details: Any) -> list[Any] | None:
+        if not details:
+            return None
+        items = details if isinstance(details, list) else [details]
+        normalized: list[Any] = []
+        for item in items:
+            if hasattr(item, "model_dump"):
+                normalized.append(item.model_dump())
+            elif isinstance(item, dict):
+                normalized.append(item)
+            else:
+                normalized.append(item)
+        return normalized or None
+
+    @staticmethod
+    def attach_reasoning_to_message(
+        message: dict[str, Any],
+        *,
+        reasoning: str = "",
+        reasoning_details: list[Any] | None = None,
+    ) -> dict[str, Any]:
+        if reasoning:
+            message["reasoning"] = reasoning
+            message["reasoning_content"] = reasoning
+        normalized = LLMClient._normalize_reasoning_details(reasoning_details)
+        if normalized:
+            message["reasoning_details"] = normalized
+        return message
+
+    async def stream_chat(
+        self,
+        messages: list[dict[str, Any]],
+        tools: list[dict[str, Any]] | None = None,
+        *,
+        model: str | None = None,
+    ) -> AsyncIterator[dict[str, Any]]:
+        use_tools = bool(tools) and self.tools_enabled
+        kwargs: dict[str, Any] = {
+            "model": model or self.model,
+            "messages": messages,
+            "stream": True,
+            "temperature": 0.7,
+        }
+        if use_tools:
+            kwargs["tools"] = tools
+        extra_body = self._reasoning_extra_body()
+        if extra_body:
+            kwargs["extra_body"] = extra_body
+
+        try:
+            stream = await self.client.chat.completions.create(**kwargs)
+        except Exception as exc:
+            logger.exception("LLM stream failed: %s", exc)
+            yield {"type": "error", "content": str(exc)}
+            yield {"type": "done", "finish_reason": "error"}
+            return
+
+        tool_calls: dict[int, dict[str, Any]] = {}
+        reasoning_parts: list[str] = []
+        reasoning_details: list[Any] = []
+
+        try:
+            async for chunk in stream:
+                if not chunk.choices:
+                    continue
+
+                choice = chunk.choices[0]
+                delta = choice.delta
+
+                if delta.content:
+                    yield {"type": "content", "content": delta.content}
+
+                reasoning_text, details = self._delta_reasoning(delta)
+                if reasoning_text:
+                    reasoning_parts.append(reasoning_text)
+                if details:
+                    reasoning_details.extend(details)
+
+                if delta.tool_calls:
+                    for tool_call in delta.tool_calls:
+                        idx = tool_call.index
+                        if idx not in tool_calls:
+                            tool_calls[idx] = {
+                                "id": tool_call.id or "",
+                                "type": "function",
+                                "function": {"name": "", "arguments": ""},
+                            }
+                        if tool_call.id:
+                            tool_calls[idx]["id"] = tool_call.id
+                        if tool_call.function:
+                            if tool_call.function.name:
+                                tool_calls[idx]["function"]["name"] = tool_call.function.name
+                            if tool_call.function.arguments:
+                                tool_calls[idx]["function"]["arguments"] += tool_call.function.arguments
+
+                usage = getattr(chunk, "usage", None)
+                if usage is not None:
+                    logger.info(
+                        "LLM stream usage: prompt=%s completion=%s total=%s",
+                        getattr(usage, "prompt_tokens", None),
+                        getattr(usage, "completion_tokens", None),
+                        getattr(usage, "total_tokens", None),
+                    )
+
+                if choice.finish_reason:
+                    reasoning = "".join(reasoning_parts)
+                    normalized_details = self._normalize_reasoning_details(reasoning_details)
+                    if reasoning or normalized_details:
+                        yield {
+                            "type": "reasoning",
+                            "reasoning": reasoning,
+                            "reasoning_details": normalized_details,
+                        }
+                    if tool_calls:
+                        yield {"type": "tool_calls", "tool_calls": list(tool_calls.values())}
+                    logger.info(
+                        "LLM stream done: model=%s finish_reason=%s tool_calls=%d "
+                        "content_in_stream=%d reasoning_len=%d",
+                        model or self.model,
+                        choice.finish_reason,
+                        len(tool_calls),
+                        len(reasoning_parts),
+                        len(reasoning),
+                    )
+                    yield {"type": "done", "finish_reason": choice.finish_reason}
+        except Exception as exc:
+            logger.exception("LLM stream read failed: %s", exc)
+            yield {"type": "error", "content": str(exc)}
+            yield {"type": "done", "finish_reason": "error"}
+
+    async def complete(
+        self,
+        messages: list[dict[str, Any]],
+        tools: list[dict[str, Any]] | None = None,
+        *,
+        temperature: float = 0.7,
+        model: str | None = None,
+        for_extraction: bool = False,
+        visible_reply: bool = False,
+    ) -> dict[str, Any]:
+        use_tools = bool(tools) and self.tools_enabled and not for_extraction
+        kwargs: dict[str, Any] = {
+            "model": model or self.model,
+            "messages": messages,
+            "temperature": temperature,
+        }
+        if use_tools:
+            kwargs["tools"] = tools
+        if for_extraction:
+            kwargs["extra_body"] = {"reasoning": {"effort": "none"}}
+        else:
+            extra_body = self._reasoning_extra_body()
+            if extra_body:
+                kwargs["extra_body"] = extra_body
+
+        response = await self.client.chat.completions.create(**kwargs)
+        usage = getattr(response, "usage", None)
+        if usage is not None:
+            logger.info(
+                "LLM complete usage: prompt=%s completion=%s total=%s model=%s",
+                getattr(usage, "prompt_tokens", None),
+                getattr(usage, "completion_tokens", None),
+                getattr(usage, "total_tokens", None),
+                kwargs.get("model"),
+            )
+        message = response.choices[0].message
+
+        content = message.content or ""
+        reasoning = ""
+        for attr in ("reasoning", "reasoning_content"):
+            value = getattr(message, attr, None)
+            if value:
+                reasoning = str(value)
+                break
+
+        if not content and reasoning and not visible_reply:
+            content = reasoning
+
+        result: dict[str, Any] = {
+            "content": content,
+            "tool_calls": [],
+            "reasoning": reasoning,
+            "reasoning_details": getattr(message, "reasoning_details", None),
+        }
+
+        if message.tool_calls:
+            result["tool_calls"] = [
+                {
+                    "id": tc.id,
+                    "type": "function",
+                    "function": {
+                        "name": tc.function.name,
+                        "arguments": tc.function.arguments,
+                    },
+                }
+                for tc in message.tool_calls
+            ]
+
+        return result
+
+    @staticmethod
+    def parse_tool_arguments(arguments: str) -> dict[str, Any]:
+        if not arguments:
+            return {}
+        try:
+            return json.loads(arguments)
+        except json.JSONDecodeError:
+            return {}
+
+    @staticmethod
+    def serialize_reasoning(
+        *,
+        reasoning: str = "",
+        reasoning_details: list[Any] | None = None,
+    ) -> str | None:
+        payload: dict[str, Any] = {}
+        if reasoning:
+            payload["reasoning"] = reasoning
+            payload["reasoning_content"] = reasoning
+        if reasoning_details:
+            payload["reasoning_details"] = reasoning_details
+        if not payload:
+            return None
+        return json.dumps(payload, ensure_ascii=False)
+
+    @staticmethod
+    def deserialize_reasoning(raw: str | None) -> dict[str, Any]:
+        if not raw:
+            return {}
+        try:
+            data = json.loads(raw)
+        except json.JSONDecodeError:
+            return {"reasoning": raw}
+        if isinstance(data, str):
+            return {"reasoning": data, "reasoning_content": data}
+        if isinstance(data, dict):
+            return data
+        return {}
+
+    async def embed(self, texts: list[str]) -> list[list[float]]:
+        settings = get_settings()
+        if not texts:
+            return []
+        response = await self.client.embeddings.create(
+            model=settings.embedding_model,
+            input=texts,
+        )
+        return [item.embedding for item in response.data]
+

backend/app/llm/client.py.refactor_bak

@@ -0,0 +1,269 @@
+import json
+import logging
+from collections.abc import AsyncIterator
+from typing import Any
+
+from openai import AsyncOpenAI
+
+from app.config import get_settings
+
+logger = logging.getLogger(__name__)
+
+
+class LLMClient:
+    def __init__(self) -> None:
+        settings = get_settings()
+        self.model = settings.openrouter_model
+        self.tools_enabled = settings.openrouter_tools_enabled
+        self.reasoning_effort = settings.openrouter_reasoning_effort.strip().lower()
+        self.client = AsyncOpenAI(
+            api_key=settings.openrouter_api_key,
+            base_url=settings.openrouter_base_url,
+        )
+
+    def _reasoning_extra_body(self) -> dict[str, Any] | None:
+        if not self.reasoning_effort:
+            return None
+        if self.reasoning_effort == "none":
+            return {"reasoning": {"effort": "none", "exclude": True}}
+        return {"reasoning": {"effort": self.reasoning_effort}}
+
+    @staticmethod
+    def _delta_reasoning(delta: Any) -> tuple[str, list[Any]]:
+        parts: list[str] = []
+        for attr in ("reasoning", "reasoning_content"):
+            value = getattr(delta, attr, None)
+            if value:
+                parts.append(str(value))
+
+        details: list[Any] = []
+        raw_details = getattr(delta, "reasoning_details", None)
+        if raw_details:
+            if isinstance(raw_details, list):
+                details.extend(raw_details)
+            else:
+                details.append(raw_details)
+
+        return "".join(parts), details
+
+    @staticmethod
+    def _normalize_reasoning_details(details: Any) -> list[Any] | None:
+        if not details:
+            return None
+        items = details if isinstance(details, list) else [details]
+        normalized: list[Any] = []
+        for item in items:
+            if hasattr(item, "model_dump"):
+                normalized.append(item.model_dump())
+            elif isinstance(item, dict):
+                normalized.append(item)
+            else:
+                normalized.append(item)
+        return normalized or None
+
+    @staticmethod
+    def attach_reasoning_to_message(
+        message: dict[str, Any],
+        *,
+        reasoning: str = "",
+        reasoning_details: list[Any] | None = None,
+    ) -> dict[str, Any]:
+        if reasoning:
+            message["reasoning"] = reasoning
+            message["reasoning_content"] = reasoning
+        normalized = LLMClient._normalize_reasoning_details(reasoning_details)
+        if normalized:
+            message["reasoning_details"] = normalized
+        return message
+
+    async def stream_chat(
+        self,
+        messages: list[dict[str, Any]],
+        tools: list[dict[str, Any]] | None = None,
+        *,
+        model: str | None = None,
+    ) -> AsyncIterator[dict[str, Any]]:
+        use_tools = bool(tools) and self.tools_enabled
+        kwargs: dict[str, Any] = {
+            "model": model or self.model,
+            "messages": messages,
+            "stream": True,
+            "temperature": 0.7,
+        }
+        if use_tools:
+            kwargs["tools"] = tools
+        extra_body = self._reasoning_extra_body()
+        if extra_body:
+            kwargs["extra_body"] = extra_body
+
+        try:
+            stream = await self.client.chat.completions.create(**kwargs)
+        except Exception as exc:
+            logger.exception("LLM stream failed: %s", exc)
+            yield {"type": "error", "content": str(exc)}
+            yield {"type": "done", "finish_reason": "error"}
+            return
+
+        tool_calls: dict[int, dict[str, Any]] = {}
+        reasoning_parts: list[str] = []
+        reasoning_details: list[Any] = []
+
+        try:
+            async for chunk in stream:
+                if not chunk.choices:
+                    continue
+
+                choice = chunk.choices[0]
+                delta = choice.delta
+
+                if delta.content:
+                    yield {"type": "content", "content": delta.content}
+
+                reasoning_text, details = self._delta_reasoning(delta)
+                if reasoning_text:
+                    reasoning_parts.append(reasoning_text)
+                if details:
+                    reasoning_details.extend(details)
+
+                if delta.tool_calls:
+                    for tool_call in delta.tool_calls:
+                        idx = tool_call.index
+                        if idx not in tool_calls:
+                            tool_calls[idx] = {
+                                "id": tool_call.id or "",
+                                "type": "function",
+                                "function": {"name": "", "arguments": ""},
+                            }
+                        if tool_call.id:
+                            tool_calls[idx]["id"] = tool_call.id
+                        if tool_call.function:
+                            if tool_call.function.name:
+                                tool_calls[idx]["function"]["name"] = tool_call.function.name
+                            if tool_call.function.arguments:
+                                tool_calls[idx]["function"]["arguments"] += tool_call.function.arguments
+
+                if choice.finish_reason:
+                    reasoning = "".join(reasoning_parts)
+                    normalized_details = self._normalize_reasoning_details(reasoning_details)
+                    if reasoning or normalized_details:
+                        yield {
+                            "type": "reasoning",
+                            "reasoning": reasoning,
+                            "reasoning_details": normalized_details,
+                        }
+                    if tool_calls:
+                        yield {"type": "tool_calls", "tool_calls": list(tool_calls.values())}
+                    logger.info(
+                        "LLM stream done: model=%s finish_reason=%s tool_calls=%d "
+                        "content_in_stream=%d reasoning_len=%d",
+                        model or self.model,
+                        choice.finish_reason,
+                        len(tool_calls),
+                        len(reasoning_parts),
+                        len(reasoning),
+                    )
+                    yield {"type": "done", "finish_reason": choice.finish_reason}
+        except Exception as exc:
+            logger.exception("LLM stream read failed: %s", exc)
+            yield {"type": "error", "content": str(exc)}
+            yield {"type": "done", "finish_reason": "error"}
+
+    async def complete(
+        self,
+        messages: list[dict[str, Any]],
+        tools: list[dict[str, Any]] | None = None,
+        *,
+        temperature: float = 0.7,
+        model: str | None = None,
+        for_extraction: bool = False,
+        visible_reply: bool = False,
+    ) -> dict[str, Any]:
+        use_tools = bool(tools) and self.tools_enabled and not for_extraction
+        kwargs: dict[str, Any] = {
+            "model": model or self.model,
+            "messages": messages,
+            "temperature": temperature,
+        }
+        if use_tools:
+            kwargs["tools"] = tools
+        if for_extraction:
+            kwargs["extra_body"] = {"reasoning": {"effort": "none"}}
+        else:
+            extra_body = self._reasoning_extra_body()
+            if extra_body:
+                kwargs["extra_body"] = extra_body
+
+        response = await self.client.chat.completions.create(**kwargs)
+        message = response.choices[0].message
+
+        content = message.content or ""
+        reasoning = ""
+        for attr in ("reasoning", "reasoning_content"):
+            value = getattr(message, attr, None)
+            if value:
+                reasoning = str(value)
+                break
+
+        if not content and reasoning and not visible_reply:
+            content = reasoning
+
+        result: dict[str, Any] = {
+            "content": content,
+            "tool_calls": [],
+            "reasoning": reasoning,
+            "reasoning_details": getattr(message, "reasoning_details", None),
+        }
+
+        if message.tool_calls:
+            result["tool_calls"] = [
+                {
+                    "id": tc.id,
+                    "type": "function",
+                    "function": {
+                        "name": tc.function.name,
+                        "arguments": tc.function.arguments,
+                    },
+                }
+                for tc in message.tool_calls
+            ]
+
+        return result
+
+    @staticmethod
+    def parse_tool_arguments(arguments: str) -> dict[str, Any]:
+        if not arguments:
+            return {}
+        try:
+            return json.loads(arguments)
+        except json.JSONDecodeError:
+            return {}
+
+    @staticmethod
+    def serialize_reasoning(
+        *,
+        reasoning: str = "",
+        reasoning_details: list[Any] | None = None,
+    ) -> str | None:
+        payload: dict[str, Any] = {}
+        if reasoning:
+            payload["reasoning"] = reasoning
+            payload["reasoning_content"] = reasoning
+        if reasoning_details:
+            payload["reasoning_details"] = reasoning_details
+        if not payload:
+            return None
+        return json.dumps(payload, ensure_ascii=False)
+
+    @staticmethod
+    def deserialize_reasoning(raw: str | None) -> dict[str, Any]:
+        if not raw:
+            return {}
+        try:
+            data = json.loads(raw)
+        except json.JSONDecodeError:
+            return {"reasoning": raw}
+        if isinstance(data, str):
+            return {"reasoning": data, "reasoning_content": data}
+        if isinstance(data, dict):
+            return data
+        return {}

backend/app/main.py

@@ -1,54 +1,65 @@
-import asyncio
-from contextlib import asynccontextmanager, suppress
-
-from fastapi import FastAPI
-from fastapi.middleware.cors import CORSMiddleware
-
-from app.api.routes import api_router
-from app.config import get_settings
-from app.db.base import init_db
-from app.fitness.watcher import fitness_watcher_loop
-from app.homelab.watcher import homelab_watcher_loop
-from app.pomodoro.watcher import pomodoro_watcher_loop
-from app.reminders.watcher import reminders_watcher_loop
-
-
-@asynccontextmanager
-async def lifespan(_: FastAPI):
-    init_db()
-    pomodoro_task = asyncio.create_task(pomodoro_watcher_loop())
-    fitness_task = asyncio.create_task(fitness_watcher_loop())
-    homelab_task = asyncio.create_task(homelab_watcher_loop())
-    reminders_task = asyncio.create_task(reminders_watcher_loop())
-    yield
-    pomodoro_task.cancel()
-    fitness_task.cancel()
-    homelab_task.cancel()
-    reminders_task.cancel()
-    with suppress(asyncio.CancelledError):
-        await pomodoro_task
-    with suppress(asyncio.CancelledError):
-        await fitness_task
-    with suppress(asyncio.CancelledError):
-        await homelab_task
-    with suppress(asyncio.CancelledError):
-        await reminders_task
-
-
-def create_app() -> FastAPI:
-    settings = get_settings()
-    app = FastAPI(title="Home AI Assistant", lifespan=lifespan)
-
-    app.add_middleware(
-        CORSMiddleware,
-        allow_origins=settings.cors_origins_list,
-        allow_credentials=True,
-        allow_methods=["*"],
-        allow_headers=["*"],
-    )
-
-    app.include_router(api_router)
-    return app
-
-
-app = create_app()
+import asyncio
+from contextlib import asynccontextmanager, suppress
+
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+
+from app.api.routes import api_router
+from app.config import get_settings
+from app.db.base import init_db
+from app.fitness.watcher import fitness_watcher_loop
+from app.homelab_scoped.watcher import homelab_watcher_loop
+from app.pomodoro.watcher import pomodoro_watcher_loop
+from app.reminders_scoped.watcher import reminders_watcher_loop
+
+
+@asynccontextmanager
+async def lifespan(_: FastAPI):
+    init_db()
+    from app.db.migrate_fitness import run_fitness_migrations
+
+    run_fitness_migrations()
+    from app.db.migrate_multi_user import run_multi_user_migrations
+
+    run_multi_user_migrations()
+    settings = get_settings()
+    if settings.rag_enabled:
+        from app.rag.store import ensure_collections
+
+        ensure_collections()
+    pomodoro_task = asyncio.create_task(pomodoro_watcher_loop())
+    fitness_task = asyncio.create_task(fitness_watcher_loop())
+    homelab_task = asyncio.create_task(homelab_watcher_loop())
+    reminders_task = asyncio.create_task(reminders_watcher_loop())
+    yield
+    pomodoro_task.cancel()
+    fitness_task.cancel()
+    homelab_task.cancel()
+    reminders_task.cancel()
+    with suppress(asyncio.CancelledError):
+        await pomodoro_task
+    with suppress(asyncio.CancelledError):
+        await fitness_task
+    with suppress(asyncio.CancelledError):
+        await homelab_task
+    with suppress(asyncio.CancelledError):
+        await reminders_task
+
+
+def create_app() -> FastAPI:
+    settings = get_settings()
+    app = FastAPI(title="Home AI Assistant", lifespan=lifespan)
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=settings.cors_origins_list,
+        allow_credentials=True,
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    app.include_router(api_router)
+    return app
+
+
+app = create_app()

backend/app/main.py.refactor_bak

@@ -0,0 +1,54 @@
+import asyncio
+from contextlib import asynccontextmanager, suppress
+
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+
+from app.api.routes import api_router
+from app.config import get_settings
+from app.db.base import init_db
+from app.fitness.watcher import fitness_watcher_loop
+from app.homelab.watcher import homelab_watcher_loop
+from app.pomodoro.watcher import pomodoro_watcher_loop
+from app.reminders.watcher import reminders_watcher_loop
+
+
+@asynccontextmanager
+async def lifespan(_: FastAPI):
+    init_db()
+    pomodoro_task = asyncio.create_task(pomodoro_watcher_loop())
+    fitness_task = asyncio.create_task(fitness_watcher_loop())
+    homelab_task = asyncio.create_task(homelab_watcher_loop())
+    reminders_task = asyncio.create_task(reminders_watcher_loop())
+    yield
+    pomodoro_task.cancel()
+    fitness_task.cancel()
+    homelab_task.cancel()
+    reminders_task.cancel()
+    with suppress(asyncio.CancelledError):
+        await pomodoro_task
+    with suppress(asyncio.CancelledError):
+        await fitness_task
+    with suppress(asyncio.CancelledError):
+        await homelab_task
+    with suppress(asyncio.CancelledError):
+        await reminders_task
+
+
+def create_app() -> FastAPI:
+    settings = get_settings()
+    app = FastAPI(title="Home AI Assistant", lifespan=lifespan)
+
+    app.add_middleware(
+        CORSMiddleware,
+        allow_origins=settings.cors_origins_list,
+        allow_credentials=True,
+        allow_methods=["*"],
+        allow_headers=["*"],
+    )
+
+    app.include_router(api_router)
+    return app
+
+
+app = create_app()

backend/app/memory/context.py

@@ -1,83 +1,89 @@
-from typing import Any
-
-from sqlalchemy.orm import Session
-
-from app.memory.service import MemoryService
-
-from app.memory.parse import is_identity_question
-
-MAX_FACTS_IN_CONTEXT = 25
-PROFILE_KEYS = ("name", "age", "timezone", "language", "notes")
-
-
-def get_memory_snapshot(db: Session, session_id: int | None = None) -> dict[str, Any]:
-    return MemoryService(db).snapshot(session_id)
-
-
-def format_memory_context(snapshot: dict[str, Any]) -> str:
-    lines = ["[Память и профиль — долгосрочный контекст]"]
-
-    profile = snapshot.get("profile") or {}
-    profile_lines = []
-    for key in PROFILE_KEYS:
-        value = (profile.get(key) or "").strip()
-        if value:
-            profile_lines.append(f"- {key}: {value}")
-    if profile_lines:
-        lines.append("Профиль пользователя:")
-        lines.extend(profile_lines)
-    else:
-        lines.append("Профиль: не заполнен (можно уточнить имя, часовой пояс).")
-
-    summary = (snapshot.get("session_summary") or "").strip()
-    if summary:
-        lines.append("")
-        lines.append("Сводка текущего чата (ранние сообщения):")
-        lines.append(summary)
-
-    facts = snapshot.get("facts") or []
-    if facts:
-        lines.append("")
-        lines.append(f"Запомненные факты ({snapshot.get('total_facts', len(facts))}):")
-        for fact in facts[:MAX_FACTS_IN_CONTEXT]:
-            lines.append(
-                f"- [{fact.get('category')}] #{fact.get('id')} {fact.get('content')}"
-            )
-    else:
-        lines.append("")
-        lines.append("Запомненные факты: пока нет.")
-
-    lines.append("")
-    lines.append(
-        "Правила памяти: "
-        "«запомни» → remember_fact (имя/возраст также пишутся в профиль). "
-        "«кто я» / «сколько мне лет» → ответь из профиля и фактов выше, БЕЗ выдумок. "
-        "Роль персонажа (сын, мать и т.п.) — стиль общения, НЕ биография пользователя. "
-        "Если профиль и факты пусты — честно скажи «не помню» и предложи запомнить. "
-        "«забудь #N» → forget_memory. "
-        "Длинный чат — update_session_summary."
-    )
-    return "\n".join(lines)
-
-
-def format_identity_hint(snapshot: dict[str, Any], user_text: str) -> str:
-    if not is_identity_question(user_text):
-        return ""
-
-    profile = snapshot.get("profile") or {}
-    facts = snapshot.get("facts") or []
-    lines = [
-        "[Вопрос об идентичности пользователя]",
-        "Ответь ТОЛЬКО из данных ниже. Не придумывай роли из сценария персонажа.",
-    ]
-    name = (profile.get("name") or "").strip()
-    age = (profile.get("age") or "").strip()
-    if name:
-        lines.append(f"Имя: {name}")
-    if age:
-        lines.append(f"Возраст: {age} лет")
-    for fact in facts:
-        lines.append(f"Факт: {fact.get('content')}")
-    if not name and not age and not facts:
-        lines.append("Данных нет — скажи, что не помнишь.")
-    return "\n".join(lines)
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.config import get_settings
+from app.memory.service import MemoryService
+
+from app.memory.parse import is_identity_question
+
+PROFILE_KEYS = ("name", "age", "timezone", "language", "notes")
+
+
+def get_memory_snapshot(
+    db: Session,
+    user_id: int,
+    session_id: int | None = None,
+    query: str | None = None,
+) -> dict[str, Any]:
+    return MemoryService(db, user_id).snapshot(session_id, query=query)
+
+
+def format_memory_context(snapshot: dict[str, Any]) -> str:
+    lines = ["[Память и профиль — долгосрочный контекст]"]
+
+    profile = snapshot.get("profile") or {}
+    profile_lines = []
+    for key in PROFILE_KEYS:
+        value = (profile.get(key) or "").strip()
+        if value:
+            profile_lines.append(f"- {key}: {value}")
+    if profile_lines:
+        lines.append("Профиль пользователя:")
+        lines.extend(profile_lines)
+    else:
+        lines.append("Профиль: не заполнен (можно уточнить имя, часовой пояс).")
+
+    summary = (snapshot.get("session_summary") or "").strip()
+    if summary:
+        lines.append("")
+        lines.append("Сводка текущего чата (ранние сообщения):")
+        lines.append(summary)
+
+    facts = snapshot.get("facts") or []
+    if facts:
+        lines.append("")
+        lines.append(f"Запомненные факты ({snapshot.get('total_facts', len(facts))}):")
+        limit = get_settings().memory_facts_in_context
+        for fact in facts[:limit]:
+            lines.append(
+                f"- [{fact.get('category')}] #{fact.get('id')} {fact.get('content')}"
+            )
+    else:
+        lines.append("")
+        lines.append("Запомненные факты: пока нет.")
+
+    lines.append("")
+    lines.append(
+        "Правила памяти: "
+        "«запомни» → remember_fact (имя/возраст также пишутся в профиль). "
+        "«кто я» / «сколько мне лет» → ответь из профиля и фактов выше, БЕЗ выдумок. "
+        "Роль персонажа (сын, мать и т.п.) — стиль общения, НЕ биография пользователя. "
+        "Если профиль и факты пусты — честно скажи «не помню» и предложи запомнить. "
+        "«забудь #N» → forget_memory. "
+        "Длинный чат — update_session_summary."
+    )
+    return "\n".join(lines)
+
+
+def format_identity_hint(snapshot: dict[str, Any], user_text: str) -> str:
+    if not is_identity_question(user_text):
+        return ""
+
+    profile = snapshot.get("profile") or {}
+    facts = snapshot.get("facts") or []
+    lines = [
+        "[Вопрос об идентичности пользователя]",
+        "Ответь ТОЛЬКО из данных ниже. Не придумывай роли из сценария персонажа.",
+    ]
+    name = (profile.get("name") or "").strip()
+    age = (profile.get("age") or "").strip()
+    if name:
+        lines.append(f"Имя: {name}")
+    if age:
+        lines.append(f"Возраст: {age} лет")
+    for fact in facts:
+        lines.append(f"Факт: {fact.get('content')}")
+    if not name and not age and not facts:
+        lines.append("Данных нет — скажи, что не помнишь.")
+    return "\n".join(lines)

backend/app/memory/context.py.refactor_bak

@@ -0,0 +1,83 @@
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.memory.service import MemoryService
+
+from app.memory.parse import is_identity_question
+
+MAX_FACTS_IN_CONTEXT = 25
+PROFILE_KEYS = ("name", "age", "timezone", "language", "notes")
+
+
+def get_memory_snapshot(db: Session, session_id: int | None = None) -> dict[str, Any]:
+    return MemoryService(db).snapshot(session_id)
+
+
+def format_memory_context(snapshot: dict[str, Any]) -> str:
+    lines = ["[Память и профиль — долгосрочный контекст]"]
+
+    profile = snapshot.get("profile") or {}
+    profile_lines = []
+    for key in PROFILE_KEYS:
+        value = (profile.get(key) or "").strip()
+        if value:
+            profile_lines.append(f"- {key}: {value}")
+    if profile_lines:
+        lines.append("Профиль пользователя:")
+        lines.extend(profile_lines)
+    else:
+        lines.append("Профиль: не заполнен (можно уточнить имя, часовой пояс).")
+
+    summary = (snapshot.get("session_summary") or "").strip()
+    if summary:
+        lines.append("")
+        lines.append("Сводка текущего чата (ранние сообщения):")
+        lines.append(summary)
+
+    facts = snapshot.get("facts") or []
+    if facts:
+        lines.append("")
+        lines.append(f"Запомненные факты ({snapshot.get('total_facts', len(facts))}):")
+        for fact in facts[:MAX_FACTS_IN_CONTEXT]:
+            lines.append(
+                f"- [{fact.get('category')}] #{fact.get('id')} {fact.get('content')}"
+            )
+    else:
+        lines.append("")
+        lines.append("Запомненные факты: пока нет.")
+
+    lines.append("")
+    lines.append(
+        "Правила памяти: "
+        "«запомни» → remember_fact (имя/возраст также пишутся в профиль). "
+        "«кто я» / «сколько мне лет» → ответь из профиля и фактов выше, БЕЗ выдумок. "
+        "Роль персонажа (сын, мать и т.п.) — стиль общения, НЕ биография пользователя. "
+        "Если профиль и факты пусты — честно скажи «не помню» и предложи запомнить. "
+        "«забудь #N» → forget_memory. "
+        "Длинный чат — update_session_summary."
+    )
+    return "\n".join(lines)
+
+
+def format_identity_hint(snapshot: dict[str, Any], user_text: str) -> str:
+    if not is_identity_question(user_text):
+        return ""
+
+    profile = snapshot.get("profile") or {}
+    facts = snapshot.get("facts") or []
+    lines = [
+        "[Вопрос об идентичности пользователя]",
+        "Ответь ТОЛЬКО из данных ниже. Не придумывай роли из сценария персонажа.",
+    ]
+    name = (profile.get("name") or "").strip()
+    age = (profile.get("age") or "").strip()
+    if name:
+        lines.append(f"Имя: {name}")
+    if age:
+        lines.append(f"Возраст: {age} лет")
+    for fact in facts:
+        lines.append(f"Факт: {fact.get('content')}")
+    if not name and not age and not facts:
+        lines.append("Данных нет — скажи, что не помнишь.")
+    return "\n".join(lines)

backend/app/memory/extract.py

@@ -1,152 +1,153 @@
-import json
-import logging
-import re
-from typing import Any
-
-from sqlalchemy.orm import Session
-
-from app.config import get_settings
-from app.llm.client import LLMClient
-from app.memory.service import MemoryService
-from app.projects.structuring import strip_markdown_json
-
-logger = logging.getLogger(__name__)
-
-SKIP_USER_PATTERN = re.compile(
-    r"^(ок|ok|да|нет|спасибо|thanks|\.{1,3}|👍|\+1)$",
-    re.IGNORECASE,
-)
-
-EXTRACTION_PROMPT = """
-Ты извлекаешь долгосрочные факты о пользователе из фрагмента диалога.
-Ответь ТОЛЬКО JSON без markdown.
-
-Схема:
-{
-  "facts": [
-    {"content": "текст факта", "category": "preference|person|habit|project|fact", "importance": 1}
-  ],
-  "profile": {"name": "", "age": "", "timezone": "", "notes": ""}
-}
-
-Правила:
-- Сохраняй устойчивое: имя, возраст, предпочтения, привычки, проекты, семья, работа.
-- НЕ сохраняй: статус помидоро, погоду, разовые команды, ролевую игру, выдумки ассистента.
-- profile — только поля с новыми значениями (пустые строки не включай).
-- facts — короткие утверждения от первого лица пользователя («люблю кофе», «меня зовут …»).
-- Если нечего сохранять — {"facts": [], "profile": {}}.
-- Не дублируй уже известное (см. текущий профиль и факты ниже).
-- importance: 5 критично (имя), 4 важно, 3 обычно, 2 мелочь.
-""".strip()
-
-
-def _should_skip_extraction(user_text: str) -> bool:
-    text = user_text.strip()
-    if len(text) < 4:
-        return True
-    if SKIP_USER_PATTERN.match(text):
-        return True
-    return False
-
-
-async def _call_extractor(
-    user_text: str,
-    assistant_text: str,
-    snapshot: dict[str, Any],
-) -> dict[str, Any]:
-    profile = snapshot.get("profile") or {}
-    facts = snapshot.get("facts") or []
-    known = [
-        f"Профиль: {json.dumps(profile, ensure_ascii=False)}",
-        "Факты:",
-        *[f"- {f.get('content')}" for f in facts[:30]],
-    ]
-
-    settings = get_settings()
-    extract_model = settings.memory_extract_model.strip() or None
-
-    llm = LLMClient()
-    result = await llm.complete(
-        [
-            {"role": "system", "content": EXTRACTION_PROMPT},
-            {
-                "role": "user",
-                "content": (
-                    "\n".join(known)
-                    + "\n\n---\nДиалог:\nПользователь: "
-                    + user_text
-                    + "\nАссистент: "
-                    + assistant_text[:1500]
-                ),
-            },
-        ],
-        temperature=0.2,
-        model=extract_model,
-        for_extraction=True,
-    )
-    raw = strip_markdown_json(result.get("content") or "")
-    if not raw:
-        return {"facts": [], "profile": {}}
-    parsed = json.loads(raw)
-    if not isinstance(parsed, dict):
-        return {"facts": [], "profile": {}}
-    return parsed
-
-
-async def extract_after_turn(
-    db: Session,
-    session_id: int,
-    user_text: str,
-    assistant_text: str,
-    *,
-    force: bool = False,
-) -> dict[str, Any]:
-    if not force and _should_skip_extraction(user_text):
-        return {"ok": True, "skipped": "short_message", "saved": []}
-
-    if not (assistant_text or "").strip():
-        return {"ok": True, "skipped": "no_assistant_reply", "saved": []}
-
-    memory = MemoryService(db)
-    snapshot = memory.snapshot(session_id)
-
-    try:
-        parsed = await _call_extractor(user_text, assistant_text, snapshot)
-    except (json.JSONDecodeError, Exception) as exc:
-        logger.warning("Memory extraction failed: %s", exc)
-        return {"ok": False, "error": str(exc), "saved": []}
-
-    saved: list[dict[str, Any]] = []
-
-    profile_updates = parsed.get("profile") or {}
-    if isinstance(profile_updates, dict):
-        filtered = {
-            k: str(v).strip()
-            for k, v in profile_updates.items()
-            if v and str(v).strip()
-        }
-        if filtered:
-            memory.update_profile(filtered)
-            saved.append({"type": "profile", "updates": filtered})
-
-    facts = parsed.get("facts") or []
-    if isinstance(facts, list):
-        for item in facts:
-            if not isinstance(item, dict):
-                continue
-            content = (item.get("content") or "").strip()
-            if not content or len(content) < 3:
-                continue
-            try:
-                result = memory.remember_fact(
-                    content,
-                    category=str(item.get("category") or "fact")[:64],
-                    importance=int(item.get("importance") or 3),
-                    session_id=session_id,
-                    source="auto",
-                )
-                saved.append({"type": "fact", **result})
-            except ValueError:
-                continue
-
-    return {"ok": True, "saved": saved, "count": len(saved)}
+import json
+import logging
+import re
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.config import get_settings
+from app.llm.client import LLMClient
+from app.memory.service import MemoryService
+from app.projects.structuring import strip_markdown_json
+
+logger = logging.getLogger(__name__)
+
+SKIP_USER_PATTERN = re.compile(
+    r"^(ок|ok|да|нет|спасибо|thanks|\.{1,3}|👍|\+1)$",
+    re.IGNORECASE,
+)
+
+EXTRACTION_PROMPT = """
+Ты извлекаешь долгосрочные факты о пользователе из фрагмента диалога.
+Ответь ТОЛЬКО JSON без markdown.
+
+Схема:
+{
+  "facts": [
+    {"content": "текст факта", "category": "preference|person|habit|project|fact", "importance": 1}
+  ],
+  "profile": {"name": "", "age": "", "timezone": "", "notes": ""}
+}
+
+Правила:
+- Сохраняй устойчивое: имя, возраст, предпочтения, привычки, проекты, семья, работа.
+- НЕ сохраняй: статус помидоро, погоду, разовые команды, ролевую игру, выдумки ассистента.
+- profile — только поля с новыми значениями (пустые строки не включай).
+- facts — короткие утверждения от первого лица пользователя («люблю кофе», «меня зовут …»).
+- Если нечего сохранять — {"facts": [], "profile": {}}.
+- Не дублируй уже известное (см. текущий профиль и факты ниже).
+- importance: 5 критично (имя), 4 важно, 3 обычно, 2 мелочь.
+""".strip()
+
+
+def _should_skip_extraction(user_text: str) -> bool:
+    text = user_text.strip()
+    if len(text) < 4:
+        return True
+    if SKIP_USER_PATTERN.match(text):
+        return True
+    return False
+
+
+async def _call_extractor(
+    user_text: str,
+    assistant_text: str,
+    snapshot: dict[str, Any],
+) -> dict[str, Any]:
+    profile = snapshot.get("profile") or {}
+    facts = snapshot.get("facts") or []
+    known = [
+        f"Профиль: {json.dumps(profile, ensure_ascii=False)}",
+        "Факты:",
+        *[f"- {f.get('content')}" for f in facts[:30]],
+    ]
+
+    settings = get_settings()
+    extract_model = settings.memory_extract_model.strip() or None
+
+    llm = LLMClient()
+    result = await llm.complete(
+        [
+            {"role": "system", "content": EXTRACTION_PROMPT},
+            {
+                "role": "user",
+                "content": (
+                    "\n".join(known)
+                    + "\n\n---\nДиалог:\nПользователь: "
+                    + user_text
+                    + "\nАссистент: "
+                    + assistant_text[:1500]
+                ),
+            },
+        ],
+        temperature=0.2,
+        model=extract_model,
+        for_extraction=True,
+    )
+    raw = strip_markdown_json(result.get("content") or "")
+    if not raw:
+        return {"facts": [], "profile": {}}
+    parsed = json.loads(raw)
+    if not isinstance(parsed, dict):
+        return {"facts": [], "profile": {}}
+    return parsed
+
+
+async def extract_after_turn(
+    db: Session,
+    session_id: int,
+    user_text: str,
+    assistant_text: str,
+    *,
+    user_id: int,
+    force: bool = False,
+) -> dict[str, Any]:
+    if not force and _should_skip_extraction(user_text):
+        return {"ok": True, "skipped": "short_message", "saved": []}
+
+    if not (assistant_text or "").strip():
+        return {"ok": True, "skipped": "no_assistant_reply", "saved": []}
+
+    memory = MemoryService(db, user_id)
+    snapshot = memory.snapshot(session_id)
+
+    try:
+        parsed = await _call_extractor(user_text, assistant_text, snapshot)
+    except (json.JSONDecodeError, Exception) as exc:
+        logger.warning("Memory extraction failed: %s", exc)
+        return {"ok": False, "error": str(exc), "saved": []}
+
+    saved: list[dict[str, Any]] = []
+
+    profile_updates = parsed.get("profile") or {}
+    if isinstance(profile_updates, dict):
+        filtered = {
+            k: str(v).strip()
+            for k, v in profile_updates.items()
+            if v and str(v).strip()
+        }
+        if filtered:
+            memory.update_profile(filtered)
+            saved.append({"type": "profile", "updates": filtered})
+
+    facts = parsed.get("facts") or []
+    if isinstance(facts, list):
+        for item in facts:
+            if not isinstance(item, dict):
+                continue
+            content = (item.get("content") or "").strip()
+            if not content or len(content) < 3:
+                continue
+            try:
+                result = memory.remember_fact(
+                    content,
+                    category=str(item.get("category") or "fact")[:64],
+                    importance=int(item.get("importance") or 3),
+                    session_id=session_id,
+                    source="auto",
+                )
+                saved.append({"type": "fact", **result})
+            except ValueError:
+                continue
+
+    return {"ok": True, "saved": saved, "count": len(saved)}

backend/app/memory/service.py

@@ -1,228 +1,300 @@
-import json
-from datetime import datetime, timezone
-from typing import Any
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.db.models import MemoryFact, SessionSummary, UserProfile
-from app.memory.parse import normalize_text, parse_identity, texts_are_similar
-
-DEFAULT_PROFILE: dict[str, Any] = {
-    "name": "",
-    "age": "",
-    "timezone": "",
-    "language": "ru",
-    "notes": "",
-}
-
-
-class MemoryService:
-    def __init__(self, db: Session):
-        self.db = db
-
-    def get_profile(self) -> dict[str, Any]:
-        row = self.db.scalar(select(UserProfile).limit(1))
-        if not row:
-            return dict(DEFAULT_PROFILE)
-        try:
-            data = json.loads(row.data_json or "{}")
-        except json.JSONDecodeError:
-            data = {}
-        merged = dict(DEFAULT_PROFILE)
-        merged.update(data)
-        return merged
-
-    def update_profile(self, updates: dict[str, Any]) -> dict[str, Any]:
-        row = self.db.scalar(select(UserProfile).limit(1))
-        if not row:
-            row = UserProfile(data_json="{}")
-            self.db.add(row)
-            self.db.flush()
-
-        current = self.get_profile()
-        for key, value in updates.items():
-            if value is None:
-                current.pop(key, None)
-            else:
-                current[key] = value
-
-        row.data_json = json.dumps(current, ensure_ascii=False)
-        row.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        return {"ok": True, "profile": current}
-
-    def _find_similar_fact(self, text: str) -> MemoryFact | None:
-        for fact in self.db.scalars(
-            select(MemoryFact).where(MemoryFact.active.is_(True))
-        ):
-            if texts_are_similar(fact.content, text):
-                return fact
-        return None
-
-    def _sync_identity_to_profile(self, text: str) -> dict[str, Any] | None:
-        parsed = parse_identity(text)
-        if not parsed:
-            return None
-        return self.update_profile(parsed)
-
-    def remember_fact(
-        self,
-        content: str,
-        *,
-        category: str = "fact",
-        source: str = "user",
-        session_id: int | None = None,
-        importance: int = 3,
-    ) -> dict[str, Any]:
-        text = content.strip()
-        if not text:
-            raise ValueError("Пустой факт")
-
-        profile_sync = self._sync_identity_to_profile(text)
-
-        existing = self._find_similar_fact(text)
-        if existing:
-            if len(text) > len(existing.content):
-                existing.content = text[:2000]
-            existing.category = category or existing.category
-            existing.importance = max(existing.importance, min(5, max(1, importance)))
-            existing.updated_at = datetime.now(timezone.utc)
-            if session_id:
-                existing.session_id = session_id
-            self.db.commit()
-            result = {
-                "ok": True,
-                "action": "updated",
-                "memory_id": existing.id,
-                "content": existing.content,
-                "category": existing.category,
-            }
-            if profile_sync:
-                result["profile"] = profile_sync.get("profile")
-            return result
-
-        fact = MemoryFact(
-            category=(category or "fact")[:64],
-            content=text[:2000],
-            source=source[:32],
-            session_id=session_id,
-            importance=min(5, max(1, importance)),
-        )
-        self.db.add(fact)
-        self.db.commit()
-        self.db.refresh(fact)
-        result = {
-            "ok": True,
-            "action": "created",
-            "memory_id": fact.id,
-            "content": fact.content,
-            "category": fact.category,
-        }
-        if profile_sync:
-            result["profile"] = profile_sync.get("profile")
-        return result
-
-    def recall_memories(
-        self,
-        *,
-        query: str | None = None,
-        category: str | None = None,
-        limit: int = 20,
-        active_only: bool = True,
-    ) -> list[dict[str, Any]]:
-        stmt = select(MemoryFact).order_by(
-            MemoryFact.importance.desc(),
-            MemoryFact.updated_at.desc(),
-        )
-        if active_only:
-            stmt = stmt.where(MemoryFact.active.is_(True))
-        if category:
-            stmt = stmt.where(MemoryFact.category == category)
-        facts = self.db.scalars(stmt.limit(100)).all()
-        if query:
-            qnorm = normalize_text(query)
-            facts = [
-                f
-                for f in facts
-                if qnorm in normalize_text(f.content)
-                or qnorm in normalize_text(f.category)
-            ]
-        facts = facts[: min(limit, 50)]
-        return [
-            {
-                "id": f.id,
-                "category": f.category,
-                "content": f.content,
-                "importance": f.importance,
-                "source": f.source,
-                "updated_at": f.updated_at.isoformat() if f.updated_at else None,
-            }
-            for f in facts
-        ]
-
-    def forget_memory(self, memory_id: int) -> dict[str, Any]:
-        fact = self.db.get(MemoryFact, memory_id)
-        if not fact:
-            raise ValueError(f"Память #{memory_id} не найдена")
-        fact.active = False
-        fact.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        return {"ok": True, "memory_id": memory_id, "forgotten": fact.content}
-
-    def get_active_facts(self, limit: int = 25) -> list[MemoryFact]:
-        return list(
-            self.db.scalars(
-                select(MemoryFact)
-                .where(MemoryFact.active.is_(True))
-                .order_by(MemoryFact.importance.desc(), MemoryFact.updated_at.desc())
-                .limit(limit)
-            ).all()
-        )
-
-    def get_session_summary(self, session_id: int) -> SessionSummary | None:
-        return self.db.scalar(
-            select(SessionSummary).where(SessionSummary.session_id == session_id)
-        )
-
-    def update_session_summary(
-        self,
-        session_id: int,
-        summary: str,
-        *,
-        message_count: int = 0,
-    ) -> dict[str, Any]:
-        text = summary.strip()
-        if not text:
-            raise ValueError("Пустая сводка")
-
-        row = self.get_session_summary(session_id)
-        if not row:
-            row = SessionSummary(session_id=session_id)
-            self.db.add(row)
-
-        row.summary = text[:4000]
-        row.message_count = message_count
-        row.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        return {"ok": True, "session_id": session_id, "summary": row.summary}
-
-    def snapshot(self, session_id: int | None = None) -> dict[str, Any]:
-        facts = self.get_active_facts()
-        summary_row = self.get_session_summary(session_id) if session_id else None
-        return {
-            "profile": self.get_profile(),
-            "facts": [
-                {
-                    "id": f.id,
-                    "category": f.category,
-                    "content": f.content,
-                    "importance": f.importance,
-                    "source": f.source,
-                    "updated_at": f.updated_at.isoformat() if f.updated_at else None,
-                }
-                for f in facts
-            ],
-            "session_summary": summary_row.summary if summary_row else "",
-            "total_facts": len(facts),
-        }
+import asyncio
+import json
+import threading
+from datetime import datetime, timezone
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import MemoryFact, SessionSummary, UserProfile
+from app.memory.parse import normalize_text, parse_identity, texts_are_similar
+
+DEFAULT_PROFILE: dict[str, Any] = {
+    "name": "",
+    "age": "",
+    "timezone": "",
+    "language": "ru",
+    "notes": "",
+}
+
+
+class MemoryService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+
+    @staticmethod
+    def _schedule_rag(coro) -> None:
+        def runner() -> None:
+            asyncio.run(coro)
+
+        threading.Thread(target=runner, daemon=True).start()
+
+    def get_profile(self) -> dict[str, Any]:
+        row = self.db.scalar(select(UserProfile).where(UserProfile.user_id == self.user_id).limit(1))
+        if not row:
+            return dict(DEFAULT_PROFILE)
+        try:
+            data = json.loads(row.data_json or "{}")
+        except json.JSONDecodeError:
+            data = {}
+        merged = dict(DEFAULT_PROFILE)
+        merged.update(data)
+        return merged
+
+    def update_profile(self, updates: dict[str, Any]) -> dict[str, Any]:
+        row = self.db.scalar(select(UserProfile).where(UserProfile.user_id == self.user_id).limit(1))
+        if not row:
+            row = UserProfile(user_id=self.user_id, data_json="{}")
+            self.db.add(row)
+            self.db.flush()
+
+        current = self.get_profile()
+        for key, value in updates.items():
+            if value is None:
+                current.pop(key, None)
+            else:
+                current[key] = value
+
+        row.data_json = json.dumps(current, ensure_ascii=False)
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return {"ok": True, "profile": current}
+
+    def _find_similar_fact(self, text: str) -> MemoryFact | None:
+        for fact in self.db.scalars(
+            select(MemoryFact).where(MemoryFact.user_id == self.user_id, MemoryFact.active.is_(True))
+        ):
+            if texts_are_similar(fact.content, text):
+                return fact
+        return None
+
+    def _sync_identity_to_profile(self, text: str) -> dict[str, Any] | None:
+        parsed = parse_identity(text)
+        if not parsed:
+            return None
+        return self.update_profile(parsed)
+
+    def remember_fact(
+        self,
+        content: str,
+        *,
+        category: str = "fact",
+        source: str = "user",
+        session_id: int | None = None,
+        importance: int = 3,
+    ) -> dict[str, Any]:
+        text = content.strip()
+        if not text:
+            raise ValueError("Пустой факт")
+
+        profile_sync = self._sync_identity_to_profile(text)
+
+        existing = self._find_similar_fact(text)
+        if existing:
+            if len(text) > len(existing.content):
+                existing.content = text[:2000]
+            existing.category = category or existing.category
+            existing.importance = max(existing.importance, min(5, max(1, importance)))
+            existing.updated_at = datetime.now(timezone.utc)
+            if session_id:
+                existing.session_id = session_id
+            self.db.commit()
+            from app.rag.ingest import index_memory_fact
+
+            self._schedule_rag(index_memory_fact(existing))
+            result = {
+                "ok": True,
+                "action": "updated",
+                "memory_id": existing.id,
+                "content": existing.content,
+                "category": existing.category,
+            }
+            if profile_sync:
+                result["profile"] = profile_sync.get("profile")
+            return result
+
+        fact = MemoryFact(
+            user_id=self.user_id,
+            category=(category or "fact")[:64],
+            content=text[:2000],
+            source=source[:32],
+            session_id=session_id,
+            importance=min(5, max(1, importance)),
+        )
+        self.db.add(fact)
+        self.db.commit()
+        self.db.refresh(fact)
+        from app.rag.ingest import index_memory_fact
+
+        self._schedule_rag(index_memory_fact(fact))
+        result = {
+            "ok": True,
+            "action": "created",
+            "memory_id": fact.id,
+            "content": fact.content,
+            "category": fact.category,
+        }
+        if profile_sync:
+            result["profile"] = profile_sync.get("profile")
+        return result
+
+    def recall_memories(
+        self,
+        *,
+        query: str | None = None,
+        category: str | None = None,
+        limit: int = 20,
+        active_only: bool = True,
+    ) -> list[dict[str, Any]]:
+        stmt = select(MemoryFact).where(MemoryFact.user_id == self.user_id).order_by(
+            MemoryFact.importance.desc(),
+            MemoryFact.updated_at.desc(),
+        )
+        if active_only:
+            stmt = stmt.where(MemoryFact.active.is_(True))
+        if category:
+            stmt = stmt.where(MemoryFact.category == category)
+        facts = self.db.scalars(stmt.limit(100)).all()
+        if query:
+            qnorm = normalize_text(query)
+            facts = [
+                f
+                for f in facts
+                if qnorm in normalize_text(f.content)
+                or qnorm in normalize_text(f.category)
+            ]
+        facts = facts[: min(limit, 50)]
+        return [
+            {
+                "id": f.id,
+                "category": f.category,
+                "content": f.content,
+                "importance": f.importance,
+                "source": f.source,
+                "updated_at": f.updated_at.isoformat() if f.updated_at else None,
+            }
+            for f in facts
+        ]
+
+    def forget_memory(self, memory_id: int) -> dict[str, Any]:
+        fact = self.db.get(MemoryFact, memory_id)
+        if not fact or fact.user_id != self.user_id:
+            raise ValueError(f"Память #{memory_id} не найдена")
+        fact.active = False
+        fact.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        from app.rag.ingest import deactivate_memory_fact
+
+        self._schedule_rag(deactivate_memory_fact(memory_id))
+        return {"ok": True, "memory_id": memory_id, "forgotten": fact.content}
+
+    def get_active_facts(self, limit: int = 25) -> list[MemoryFact]:
+        return list(
+            self.db.scalars(
+                select(MemoryFact)
+                .where(MemoryFact.user_id == self.user_id, MemoryFact.active.is_(True))
+                .order_by(MemoryFact.importance.desc(), MemoryFact.updated_at.desc())
+                .limit(limit)
+            ).all()
+        )
+
+    def get_session_summary(self, session_id: int) -> SessionSummary | None:
+        from app.db.models import ChatSession
+
+        session = self.db.get(ChatSession, session_id)
+        if not session or session.user_id != self.user_id:
+            return None
+        return self.db.scalar(
+            select(SessionSummary).where(SessionSummary.session_id == session_id)
+        )
+
+    def update_session_summary(
+        self,
+        session_id: int,
+        summary: str,
+        *,
+        message_count: int = 0,
+    ) -> dict[str, Any]:
+        text = summary.strip()
+        if not text:
+            raise ValueError("Пустая сводка")
+
+        from app.db.models import ChatSession
+
+        session = self.db.get(ChatSession, session_id)
+        if not session or session.user_id != self.user_id:
+            raise ValueError("Session not found")
+
+        row = self.db.scalar(
+            select(SessionSummary).where(SessionSummary.session_id == session_id)
+        )
+        if not row:
+            row = SessionSummary(session_id=session_id)
+            self.db.add(row)
+
+        row.summary = text[:4000]
+        row.message_count = message_count
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        from app.rag.ingest import index_session_summary
+
+        self._schedule_rag(index_session_summary(session_id, row.summary))
+        return {"ok": True, "session_id": session_id, "summary": row.summary}
+
+    def snapshot(self, session_id: int | None = None, query: str | None = None) -> dict[str, Any]:
+        from app.config import get_settings
+        from app.settings.service import SettingsService
+
+        settings = get_settings()
+        svc = SettingsService(self.db)
+        rag_on = bool(svc.get_effective("rag_enabled")) and settings.rag_enabled
+        facts_payload: list[dict[str, Any]]
+        total_facts = len(self.get_active_facts(limit=500))
+        if rag_on and (query or "").strip():
+            async def _load() -> list[dict[str, Any]]:
+                from app.rag.retriever import retrieve_memory_facts
+
+                top_k = int(svc.get_effective("rag_top_k"))
+                return await retrieve_memory_facts(query or "", user_id=self.user_id, top_k=top_k)
+
+            try:
+                rag_facts = asyncio.run(_load())
+            except Exception:
+                rag_facts = []
+            if rag_facts:
+                facts_payload = rag_facts
+            else:
+                facts = self.get_active_facts(limit=settings.memory_facts_in_context)
+                facts_payload = [
+                    {
+                        "id": f.id,
+                        "category": f.category,
+                        "content": f.content,
+                        "importance": f.importance,
+                        "source": f.source,
+                        "updated_at": f.updated_at.isoformat() if f.updated_at else None,
+                    }
+                    for f in facts
+                ]
+        else:
+            facts = self.get_active_facts(limit=settings.memory_facts_in_context)
+            facts_payload = [
+                {
+                    "id": f.id,
+                    "category": f.category,
+                    "content": f.content,
+                    "importance": f.importance,
+                    "source": f.source,
+                    "updated_at": f.updated_at.isoformat() if f.updated_at else None,
+                }
+                for f in facts
+            ]
+        summary_row = self.get_session_summary(session_id) if session_id else None
+        return {
+            "profile": self.get_profile(),
+            "facts": facts_payload,
+            "session_summary": summary_row.summary if summary_row else "",
+            "total_facts": total_facts,
+        }

backend/app/memory/service.py.refactor_bak

@@ -0,0 +1,228 @@
+import json
+from datetime import datetime, timezone
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import MemoryFact, SessionSummary, UserProfile
+from app.memory.parse import normalize_text, parse_identity, texts_are_similar
+
+DEFAULT_PROFILE: dict[str, Any] = {
+    "name": "",
+    "age": "",
+    "timezone": "",
+    "language": "ru",
+    "notes": "",
+}
+
+
+class MemoryService:
+    def __init__(self, db: Session):
+        self.db = db
+
+    def get_profile(self) -> dict[str, Any]:
+        row = self.db.scalar(select(UserProfile).limit(1))
+        if not row:
+            return dict(DEFAULT_PROFILE)
+        try:
+            data = json.loads(row.data_json or "{}")
+        except json.JSONDecodeError:
+            data = {}
+        merged = dict(DEFAULT_PROFILE)
+        merged.update(data)
+        return merged
+
+    def update_profile(self, updates: dict[str, Any]) -> dict[str, Any]:
+        row = self.db.scalar(select(UserProfile).limit(1))
+        if not row:
+            row = UserProfile(data_json="{}")
+            self.db.add(row)
+            self.db.flush()
+
+        current = self.get_profile()
+        for key, value in updates.items():
+            if value is None:
+                current.pop(key, None)
+            else:
+                current[key] = value
+
+        row.data_json = json.dumps(current, ensure_ascii=False)
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return {"ok": True, "profile": current}
+
+    def _find_similar_fact(self, text: str) -> MemoryFact | None:
+        for fact in self.db.scalars(
+            select(MemoryFact).where(MemoryFact.active.is_(True))
+        ):
+            if texts_are_similar(fact.content, text):
+                return fact
+        return None
+
+    def _sync_identity_to_profile(self, text: str) -> dict[str, Any] | None:
+        parsed = parse_identity(text)
+        if not parsed:
+            return None
+        return self.update_profile(parsed)
+
+    def remember_fact(
+        self,
+        content: str,
+        *,
+        category: str = "fact",
+        source: str = "user",
+        session_id: int | None = None,
+        importance: int = 3,
+    ) -> dict[str, Any]:
+        text = content.strip()
+        if not text:
+            raise ValueError("Пустой факт")
+
+        profile_sync = self._sync_identity_to_profile(text)
+
+        existing = self._find_similar_fact(text)
+        if existing:
+            if len(text) > len(existing.content):
+                existing.content = text[:2000]
+            existing.category = category or existing.category
+            existing.importance = max(existing.importance, min(5, max(1, importance)))
+            existing.updated_at = datetime.now(timezone.utc)
+            if session_id:
+                existing.session_id = session_id
+            self.db.commit()
+            result = {
+                "ok": True,
+                "action": "updated",
+                "memory_id": existing.id,
+                "content": existing.content,
+                "category": existing.category,
+            }
+            if profile_sync:
+                result["profile"] = profile_sync.get("profile")
+            return result
+
+        fact = MemoryFact(
+            category=(category or "fact")[:64],
+            content=text[:2000],
+            source=source[:32],
+            session_id=session_id,
+            importance=min(5, max(1, importance)),
+        )
+        self.db.add(fact)
+        self.db.commit()
+        self.db.refresh(fact)
+        result = {
+            "ok": True,
+            "action": "created",
+            "memory_id": fact.id,
+            "content": fact.content,
+            "category": fact.category,
+        }
+        if profile_sync:
+            result["profile"] = profile_sync.get("profile")
+        return result
+
+    def recall_memories(
+        self,
+        *,
+        query: str | None = None,
+        category: str | None = None,
+        limit: int = 20,
+        active_only: bool = True,
+    ) -> list[dict[str, Any]]:
+        stmt = select(MemoryFact).order_by(
+            MemoryFact.importance.desc(),
+            MemoryFact.updated_at.desc(),
+        )
+        if active_only:
+            stmt = stmt.where(MemoryFact.active.is_(True))
+        if category:
+            stmt = stmt.where(MemoryFact.category == category)
+        facts = self.db.scalars(stmt.limit(100)).all()
+        if query:
+            qnorm = normalize_text(query)
+            facts = [
+                f
+                for f in facts
+                if qnorm in normalize_text(f.content)
+                or qnorm in normalize_text(f.category)
+            ]
+        facts = facts[: min(limit, 50)]
+        return [
+            {
+                "id": f.id,
+                "category": f.category,
+                "content": f.content,
+                "importance": f.importance,
+                "source": f.source,
+                "updated_at": f.updated_at.isoformat() if f.updated_at else None,
+            }
+            for f in facts
+        ]
+
+    def forget_memory(self, memory_id: int) -> dict[str, Any]:
+        fact = self.db.get(MemoryFact, memory_id)
+        if not fact:
+            raise ValueError(f"Память #{memory_id} не найдена")
+        fact.active = False
+        fact.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return {"ok": True, "memory_id": memory_id, "forgotten": fact.content}
+
+    def get_active_facts(self, limit: int = 25) -> list[MemoryFact]:
+        return list(
+            self.db.scalars(
+                select(MemoryFact)
+                .where(MemoryFact.active.is_(True))
+                .order_by(MemoryFact.importance.desc(), MemoryFact.updated_at.desc())
+                .limit(limit)
+            ).all()
+        )
+
+    def get_session_summary(self, session_id: int) -> SessionSummary | None:
+        return self.db.scalar(
+            select(SessionSummary).where(SessionSummary.session_id == session_id)
+        )
+
+    def update_session_summary(
+        self,
+        session_id: int,
+        summary: str,
+        *,
+        message_count: int = 0,
+    ) -> dict[str, Any]:
+        text = summary.strip()
+        if not text:
+            raise ValueError("Пустая сводка")
+
+        row = self.get_session_summary(session_id)
+        if not row:
+            row = SessionSummary(session_id=session_id)
+            self.db.add(row)
+
+        row.summary = text[:4000]
+        row.message_count = message_count
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return {"ok": True, "session_id": session_id, "summary": row.summary}
+
+    def snapshot(self, session_id: int | None = None) -> dict[str, Any]:
+        facts = self.get_active_facts()
+        summary_row = self.get_session_summary(session_id) if session_id else None
+        return {
+            "profile": self.get_profile(),
+            "facts": [
+                {
+                    "id": f.id,
+                    "category": f.category,
+                    "content": f.content,
+                    "importance": f.importance,
+                    "source": f.source,
+                    "updated_at": f.updated_at.isoformat() if f.updated_at else None,
+                }
+                for f in facts
+            ],
+            "session_summary": summary_row.summary if summary_row else "",
+            "total_facts": len(facts),
+        }

backend/app/pomodoro/completion.py

@@ -1,91 +1,92 @@
-import logging
-
-from sqlalchemy.orm import Session
-
-from app.character.service import CharacterService
-from app.chat.notice_inbox import post_character_comment_to_latest_chat, post_notice_to_latest_chat
-from app.chat.notices import format_phase_completed_notice
-from app.db.models import PomodoroSession
-from app.llm.client import LLMClient
-from app.pomodoro.cycle import PHASE_LONG_BREAK, PHASE_SHORT_BREAK, PHASE_WORK, CycleManager
-from app.pomodoro.service import PomodoroService
-
-logger = logging.getLogger(__name__)
-
-PHASE_LABELS = {
-    PHASE_WORK: "работа",
-    PHASE_SHORT_BREAK: "короткий перерыв",
-    PHASE_LONG_BREAK: "длинный перерыв",
-}
-
-
-class PomodoroCompletionHandler:
-    def __init__(self, db: Session):
-        self.db = db
-        self.pomodoro = PomodoroService(db)
-        self.cycle = CycleManager(db)
-        self.llm = LLMClient()
-        self.character = CharacterService()
-
-    async def _generate_llm_comment(
-        self,
-        session: PomodoroSession,
-        next_phase: str | None,
-    ) -> str:
-        cycle = self.cycle.to_dict()
-        phase_label = PHASE_LABELS.get(session.phase, session.phase)
-        next_label = PHASE_LABELS.get(next_phase, "пауза") if next_phase else "отдых, цикл сброшен"
-        work_done = cycle["completed_work_sessions"]
-        if session.phase == PHASE_WORK:
-            work_done += 1
-
-        system = self.character.get_system_prompt()
-        user_prompt = f"""Фаза помидоро «{phase_label}» только что завершилась.
-Задача: {session.task_note or 'без описания'}
-Прогресс цикла: {work_done}/{cycle['sessions_until_long_break']} работ.
-Следующая фаза: {next_label}.
-
-Напиши пользователю короткое сообщение (2-4 предложения) на русском: поздравь, поддержи или предложи отдохнуть. Без markdown и без эмодзи."""
-
-        result = await self.llm.complete(
-            [
-                {"role": "system", "content": system},
-                {"role": "user", "content": user_prompt},
-            ],
-            temperature=0.8,
-            visible_reply=True,
-        )
-        return (result.get("content") or "").strip() or "Фаза завершена. Хорошая работа."
-
-    def _resolve_next_phase(self, session: PomodoroSession) -> str | None:
-        phase = session.phase
-        cycle = self.cycle.get()
-        if phase == PHASE_WORK:
-            if cycle.completed_work_sessions + 1 >= cycle.sessions_until_long_break:
-                return PHASE_LONG_BREAK
-            return PHASE_SHORT_BREAK
-        if phase == PHASE_SHORT_BREAK:
-            return PHASE_WORK
-        if phase == PHASE_LONG_BREAK:
-            return None
-        return None
-
-    async def process(self, session: PomodoroSession) -> None:
-        if session.completion_notified:
-            return
-
-        next_phase = self._resolve_next_phase(session)
-        notice = format_phase_completed_notice(session, next_phase)
-        post_notice_to_latest_chat(notice)
-
-        try:
-            comment = await self._generate_llm_comment(session, next_phase)
-            if comment:
-                post_character_comment_to_latest_chat(comment)
-        except Exception:
-            logger.exception("Pomodoro LLM comment failed (phase=%s)", session.phase)
-
-        self.cycle.bump_notify_seq()
-        self.pomodoro.mark_notified(session)
-        self.pomodoro.advance_after_completion(session)
-        logger.info("Pomodoro phase completed (phase=%s, next=%s)", session.phase, next_phase)
+import logging
+
+from sqlalchemy.orm import Session
+
+from app.character.service import CharacterService
+from app.chat.notice_inbox import post_character_comment_to_latest_chat, post_notice_to_latest_chat
+from app.chat.notices import format_phase_completed_notice
+from app.db.models import PomodoroSession
+from app.llm.client import LLMClient
+from app.pomodoro.cycle import PHASE_LONG_BREAK, PHASE_SHORT_BREAK, PHASE_WORK, CycleManager
+from app.pomodoro.service import PomodoroService
+
+logger = logging.getLogger(__name__)
+
+PHASE_LABELS = {
+    PHASE_WORK: "работа",
+    PHASE_SHORT_BREAK: "короткий перерыв",
+    PHASE_LONG_BREAK: "длинный перерыв",
+}
+
+
+class PomodoroCompletionHandler:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+        self.pomodoro = PomodoroService(db, user_id)
+        self.cycle = CycleManager(db, user_id)
+        self.llm = LLMClient()
+        self.character = CharacterService(db, user_id)
+
+    async def _generate_llm_comment(
+        self,
+        session: PomodoroSession,
+        next_phase: str | None,
+    ) -> str:
+        cycle = self.cycle.to_dict()
+        phase_label = PHASE_LABELS.get(session.phase, session.phase)
+        next_label = PHASE_LABELS.get(next_phase, "пауза") if next_phase else "отдых, цикл сброшен"
+        work_done = cycle["completed_work_sessions"]
+        if session.phase == PHASE_WORK:
+            work_done += 1
+
+        system = self.character.get_system_prompt()
+        user_prompt = f"""Фаза помидоро «{phase_label}» только что завершилась.
+Задача: {session.task_note or 'без описания'}
+Прогресс цикла: {work_done}/{cycle['sessions_until_long_break']} работ.
+Следующая фаза: {next_label}.
+
+Напиши пользователю короткое сообщение (2-4 предложения) на русском: поздравь, поддержи или предложи отдохнуть. Без markdown и без эмодзи."""
+
+        result = await self.llm.complete(
+            [
+                {"role": "system", "content": system},
+                {"role": "user", "content": user_prompt},
+            ],
+            temperature=0.8,
+            visible_reply=True,
+        )
+        return (result.get("content") or "").strip() or "Фаза завершена. Хорошая работа."
+
+    def _resolve_next_phase(self, session: PomodoroSession) -> str | None:
+        phase = session.phase
+        cycle = self.cycle.get()
+        if phase == PHASE_WORK:
+            if cycle.completed_work_sessions + 1 >= cycle.sessions_until_long_break:
+                return PHASE_LONG_BREAK
+            return PHASE_SHORT_BREAK
+        if phase == PHASE_SHORT_BREAK:
+            return PHASE_WORK
+        if phase == PHASE_LONG_BREAK:
+            return None
+        return None
+
+    async def process(self, session: PomodoroSession) -> None:
+        if session.completion_notified:
+            return
+
+        next_phase = self._resolve_next_phase(session)
+        notice = format_phase_completed_notice(session, next_phase)
+        post_notice_to_latest_chat(notice, self.user_id)
+
+        try:
+            comment = await self._generate_llm_comment(session, next_phase)
+            if comment:
+                post_character_comment_to_latest_chat(comment, self.user_id)
+        except Exception:
+            logger.exception("Pomodoro LLM comment failed (phase=%s)", session.phase)
+
+        self.cycle.bump_notify_seq()
+        self.pomodoro.mark_notified(session)
+        self.pomodoro.advance_after_completion(session)
+        logger.info("Pomodoro phase completed (phase=%s, next=%s)", session.phase, next_phase)

backend/app/pomodoro/cycle.py

@@ -1,89 +1,90 @@
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.db.models import PomodoroCycle
-
-PHASE_WORK = "work"
-PHASE_SHORT_BREAK = "short_break"
-PHASE_LONG_BREAK = "long_break"
-
-
-class CycleManager:
-    def __init__(self, db: Session):
-        self.db = db
-
-    def get(self) -> PomodoroCycle:
-        cycle = self.db.scalar(select(PomodoroCycle).limit(1))
-        if not cycle:
-            cycle = PomodoroCycle()
-            self.db.add(cycle)
-            self.db.commit()
-            self.db.refresh(cycle)
-        return cycle
-
-    def to_dict(self, cycle: PomodoroCycle | None = None) -> dict:
-        c = cycle or self.get()
-        return {
-            "completed_work_sessions": c.completed_work_sessions,
-            "sessions_until_long_break": c.sessions_until_long_break,
-            "task_note": c.task_note,
-            "work_duration_min": c.work_duration_min,
-            "short_break_min": c.short_break_min,
-            "long_break_min": c.long_break_min,
-            "auto_advance": c.auto_advance,
-            "chat_notify_seq": c.chat_notify_seq,
-        }
-
-    def reset(self, clear_task: bool = False) -> dict:
-        cycle = self.get()
-        cycle.completed_work_sessions = 0
-        if clear_task:
-            cycle.task_note = ""
-        self.db.commit()
-        self.db.refresh(cycle)
-        return self.to_dict(cycle)
-
-    def bump_notify_seq(self) -> int:
-        cycle = self.get()
-        cycle.chat_notify_seq += 1
-        self.db.commit()
-        self.db.refresh(cycle)
-        return cycle.chat_notify_seq
-
-    def on_work_completed(self) -> str:
-        """Returns next phase: short_break or long_break."""
-        cycle = self.get()
-        cycle.completed_work_sessions += 1
-        if cycle.completed_work_sessions >= cycle.sessions_until_long_break:
-            next_phase = PHASE_LONG_BREAK
-        else:
-            next_phase = PHASE_SHORT_BREAK
-        self.db.commit()
-        return next_phase
-
-    def on_long_break_completed(self) -> None:
-        cycle = self.get()
-        cycle.completed_work_sessions = 0
-        self.db.commit()
-
-    def duration_for_phase(self, phase: str, cycle: PomodoroCycle | None = None) -> int:
-        c = cycle or self.get()
-        if phase == PHASE_WORK:
-            return c.work_duration_min
-        if phase == PHASE_SHORT_BREAK:
-            return c.short_break_min
-        if phase == PHASE_LONG_BREAK:
-            return c.long_break_min
-        return c.work_duration_min
-
-    def next_phase_after(self, completed_phase: str) -> str | None:
-        if completed_phase == PHASE_WORK:
-            cycle = self.get()
-            if cycle.completed_work_sessions >= cycle.sessions_until_long_break:
-                return PHASE_LONG_BREAK
-            return PHASE_SHORT_BREAK
-        if completed_phase == PHASE_SHORT_BREAK:
-            return PHASE_WORK
-        if completed_phase == PHASE_LONG_BREAK:
-            return None
-        return None
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import PomodoroCycle
+
+PHASE_WORK = "work"
+PHASE_SHORT_BREAK = "short_break"
+PHASE_LONG_BREAK = "long_break"
+
+
+class CycleManager:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+
+    def get(self) -> PomodoroCycle:
+        cycle = self.db.scalar(select(PomodoroCycle).where(PomodoroCycle.user_id == self.user_id).limit(1))
+        if not cycle:
+            cycle = PomodoroCycle(user_id=self.user_id)
+            self.db.add(cycle)
+            self.db.commit()
+            self.db.refresh(cycle)
+        return cycle
+
+    def to_dict(self, cycle: PomodoroCycle | None = None) -> dict:
+        c = cycle or self.get()
+        return {
+            "completed_work_sessions": c.completed_work_sessions,
+            "sessions_until_long_break": c.sessions_until_long_break,
+            "task_note": c.task_note,
+            "work_duration_min": c.work_duration_min,
+            "short_break_min": c.short_break_min,
+            "long_break_min": c.long_break_min,
+            "auto_advance": c.auto_advance,
+            "chat_notify_seq": c.chat_notify_seq,
+        }
+
+    def reset(self, clear_task: bool = False) -> dict:
+        cycle = self.get()
+        cycle.completed_work_sessions = 0
+        if clear_task:
+            cycle.task_note = ""
+        self.db.commit()
+        self.db.refresh(cycle)
+        return self.to_dict(cycle)
+
+    def bump_notify_seq(self) -> int:
+        cycle = self.get()
+        cycle.chat_notify_seq += 1
+        self.db.commit()
+        self.db.refresh(cycle)
+        return cycle.chat_notify_seq
+
+    def on_work_completed(self) -> str:
+        """Returns next phase: short_break or long_break."""
+        cycle = self.get()
+        cycle.completed_work_sessions += 1
+        if cycle.completed_work_sessions >= cycle.sessions_until_long_break:
+            next_phase = PHASE_LONG_BREAK
+        else:
+            next_phase = PHASE_SHORT_BREAK
+        self.db.commit()
+        return next_phase
+
+    def on_long_break_completed(self) -> None:
+        cycle = self.get()
+        cycle.completed_work_sessions = 0
+        self.db.commit()
+
+    def duration_for_phase(self, phase: str, cycle: PomodoroCycle | None = None) -> int:
+        c = cycle or self.get()
+        if phase == PHASE_WORK:
+            return c.work_duration_min
+        if phase == PHASE_SHORT_BREAK:
+            return c.short_break_min
+        if phase == PHASE_LONG_BREAK:
+            return c.long_break_min
+        return c.work_duration_min
+
+    def next_phase_after(self, completed_phase: str) -> str | None:
+        if completed_phase == PHASE_WORK:
+            cycle = self.get()
+            if cycle.completed_work_sessions >= cycle.sessions_until_long_break:
+                return PHASE_LONG_BREAK
+            return PHASE_SHORT_BREAK
+        if completed_phase == PHASE_SHORT_BREAK:
+            return PHASE_WORK
+        if completed_phase == PHASE_LONG_BREAK:
+            return None
+        return None

backend/app/pomodoro/service.py

@@ -1,287 +1,296 @@
-from datetime import datetime, timezone
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.db.models import PomodoroSession
-from app.pomodoro.cycle import (
-    PHASE_LONG_BREAK,
-    PHASE_SHORT_BREAK,
-    PHASE_WORK,
-    CycleManager,
-)
-
-
-def _utcnow() -> datetime:
-    return datetime.now(timezone.utc)
-
-
-class PomodoroService:
-    def __init__(self, db: Session):
-        self.db = db
-        self.cycle = CycleManager(db)
-
-    def _get_active(self) -> PomodoroSession | None:
-        stmt = (
-            select(PomodoroSession)
-            .where(PomodoroSession.status.in_(("running", "paused")))
-            .order_by(PomodoroSession.id.desc())
-            .limit(1)
-        )
-        return self.db.scalar(stmt)
-
-    def _elapsed(self, session: PomodoroSession) -> int:
-        elapsed = session.elapsed_seconds
-        if session.status == "running" and session.started_at:
-            started = session.started_at
-            if started.tzinfo is None:
-                started = started.replace(tzinfo=timezone.utc)
-            delta = _utcnow() - started
-            elapsed += int(delta.total_seconds())
-        return elapsed
-
-    def _remaining(self, session: PomodoroSession) -> int:
-        total = session.duration_min * 60
-        return max(0, total - self._elapsed(session))
-
-    def _try_auto_complete(self, session: PomodoroSession) -> bool:
-        if session.status != "running":
-            return False
-        if self._remaining(session) > 0:
-            return False
-        self._finalize_session(session, auto=True)
-        return True
-
-    def _finalize_session(
-        self,
-        session: PomodoroSession,
-        *,
-        auto: bool,
-        result: str = "",
-        completed: bool | None = None,
-        cancelled: bool = False,
-    ) -> None:
-        session.elapsed_seconds = self._elapsed(session)
-        session.started_at = None
-        session.finished_at = _utcnow()
-        session.completion_notified = False
-        session.result = result or None
-
-        if cancelled:
-            session.status = "cancelled"
-            session.completed = False
-        elif completed is not None:
-            session.status = "completed"
-            session.completed = completed
-        else:
-            session.status = "completed"
-            session.completed = True
-
-        self.db.commit()
-        self.db.refresh(session)
-
-    def _start_phase(
-        self,
-        phase: str,
-        *,
-        duration_min: int | None = None,
-        task_note: str | None = None,
-    ) -> PomodoroSession:
-        active = self._get_active()
-        if active:
-            raise ValueError("Таймер уже запущен. Сначала остановите текущую сессию.")
-
-        cycle = self.cycle.get()
-        if task_note is not None:
-            cycle.task_note = task_note
-        elif phase == PHASE_WORK and not cycle.task_note:
-            cycle.task_note = ""
-
-        duration = duration_min or self.cycle.duration_for_phase(phase, cycle)
-        note = task_note if task_note is not None else cycle.task_note
-
-        session = PomodoroSession(
-            status="running",
-            phase=phase,
-            duration_min=duration,
-            task_note=note,
-            started_at=_utcnow(),
-        )
-        self.db.add(session)
-        self.db.commit()
-        self.db.refresh(session)
-        return session
-
-    def _to_status_dict(self, session: PomodoroSession | None) -> dict:
-        cycle_dict = self.cycle.to_dict()
-        if not session:
-            return {
-                "status": "idle",
-                "phase": PHASE_WORK,
-                "duration_min": cycle_dict["work_duration_min"],
-                "task_note": cycle_dict["task_note"],
-                "elapsed_seconds": 0,
-                "remaining_seconds": 0,
-                "session_id": None,
-                "cycle": cycle_dict,
-            }
-
-        elapsed = self._elapsed(session)
-        total = session.duration_min * 60
-        remaining = max(0, total - elapsed)
-
-        return {
-            "status": session.status,
-            "phase": session.phase,
-            "duration_min": session.duration_min,
-            "task_note": session.task_note,
-            "elapsed_seconds": elapsed,
-            "remaining_seconds": remaining,
-            "session_id": session.id,
-            "started_at": session.started_at.isoformat() if session.started_at else None,
-            "finished_at": session.finished_at.isoformat() if session.finished_at else None,
-            "cycle": cycle_dict,
-        }
-
-    def get_status(self) -> dict:
-        active = self._get_active()
-        if active:
-            self._try_auto_complete(active)
-            active = self._get_active()
-        return self._to_status_dict(active)
-
-    def start_work(self, duration_min: int | None = None, task_note: str = "") -> dict:
-        session = self._start_phase(
-            PHASE_WORK,
-            duration_min=duration_min,
-            task_note=task_note,
-        )
-        return self._to_status_dict(session)
-
-    def start_short_break(self, duration_min: int | None = None) -> dict:
-        session = self._start_phase(PHASE_SHORT_BREAK, duration_min=duration_min)
-        return self._to_status_dict(session)
-
-    def start_long_break(self, duration_min: int | None = None) -> dict:
-        session = self._start_phase(PHASE_LONG_BREAK, duration_min=duration_min)
-        return self._to_status_dict(session)
-
-    def start(self, duration_min: int = 25, task_note: str = "") -> dict:
-        return self.start_work(duration_min=duration_min, task_note=task_note)
-
-    def pause(self) -> dict:
-        session = self._get_active()
-        if not session or session.status != "running":
-            raise ValueError("Нет активного запущенного таймера.")
-
-        session.elapsed_seconds = self._elapsed(session)
-        session.status = "paused"
-        session.paused_at = _utcnow()
-        session.started_at = None
-        self.db.commit()
-        self.db.refresh(session)
-        return self._to_status_dict(session)
-
-    def resume(self) -> dict:
-        session = self._get_active()
-        if not session or session.status != "paused":
-            raise ValueError("Нет таймера на паузе.")
-
-        session.status = "running"
-        session.started_at = _utcnow()
-        session.paused_at = None
-        self.db.commit()
-        self.db.refresh(session)
-        return self._to_status_dict(session)
-
-    def stop(self, result: str = "", completed: bool = False) -> dict:
-        session = self._get_active()
-        if not session:
-            raise ValueError("Нет активного таймера.")
-
-        if completed:
-            self._finalize_session(session, auto=False, result=result, completed=True)
-        else:
-            self._finalize_session(session, auto=False, result=result, cancelled=True)
-            session.completion_notified = True
-            self.db.commit()
-        return self._to_status_dict(None)
-
-    def reset_cycle(self, clear_task: bool = False) -> dict:
-        active = self._get_active()
-        if active:
-            self._finalize_session(active, auto=False, cancelled=True)
-            active.completion_notified = True
-            self.db.commit()
-        cycle = self.cycle.reset(clear_task=clear_task)
-        status = self._to_status_dict(None)
-        status["cycle"] = cycle
-        return status
-
-    def skip_phase(self) -> dict:
-        session = self._get_active()
-        if not session:
-            raise ValueError("Нет активного таймера.")
-
-        self._finalize_session(session, auto=True)
-        return self._to_status_dict(None)
-
-    def get_pending_completions(self) -> list[PomodoroSession]:
-        stmt = (
-            select(PomodoroSession)
-            .where(
-                PomodoroSession.status == "completed",
-                PomodoroSession.completed.is_(True),
-                PomodoroSession.completion_notified.is_(False),
-            )
-            .order_by(PomodoroSession.id.asc())
-        )
-        return list(self.db.scalars(stmt))
-
-    def mark_notified(self, session: PomodoroSession) -> None:
-        session.completion_notified = True
-        self.db.commit()
-
-    def advance_after_completion(self, session: PomodoroSession) -> dict | None:
-        """Update cycle counters and auto-start next phase. Returns new status or None."""
-        phase = session.phase
-        cycle = self.cycle.get()
-
-        if phase == PHASE_WORK:
-            next_phase = self.cycle.on_work_completed()
-        elif phase == PHASE_SHORT_BREAK:
-            next_phase = PHASE_WORK
-        elif phase == PHASE_LONG_BREAK:
-            self.cycle.on_long_break_completed()
-            next_phase = None
-        else:
-            next_phase = None
-
-        if not cycle.auto_advance or next_phase is None:
-            return None
-
-        new_session = self._start_phase(next_phase)
-        return self._to_status_dict(new_session)
-
-    def history(self, limit: int = 20) -> list[dict]:
-        stmt = (
-            select(PomodoroSession)
-            .where(PomodoroSession.status.in_(("completed", "cancelled")))
-            .order_by(PomodoroSession.finished_at.desc())
-            .limit(limit)
-        )
-        sessions = self.db.scalars(stmt).all()
-        return [
-            {
-                "id": s.id,
-                "status": s.status,
-                "phase": s.phase,
-                "duration_min": s.duration_min,
-                "task_note": s.task_note,
-                "result": s.result,
-                "completed": s.completed,
-                "elapsed_seconds": s.elapsed_seconds,
-                "finished_at": s.finished_at.isoformat() if s.finished_at else None,
-            }
-            for s in sessions
-        ]
+from datetime import datetime, timezone
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import PomodoroSession
+from app.pomodoro.cycle import (
+    PHASE_LONG_BREAK,
+    PHASE_SHORT_BREAK,
+    PHASE_WORK,
+    CycleManager,
+)
+
+
+def _utcnow() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+class PomodoroService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+        self.cycle = CycleManager(db, user_id)
+
+    def _get_active(self) -> PomodoroSession | None:
+        stmt = (
+            select(PomodoroSession)
+            .where(
+                PomodoroSession.user_id == self.user_id,
+                PomodoroSession.status.in_(("running", "paused")),
+            )
+            .order_by(PomodoroSession.id.desc())
+            .limit(1)
+        )
+        return self.db.scalar(stmt)
+
+    def _elapsed(self, session: PomodoroSession) -> int:
+        elapsed = session.elapsed_seconds
+        if session.status == "running" and session.started_at:
+            started = session.started_at
+            if started.tzinfo is None:
+                started = started.replace(tzinfo=timezone.utc)
+            delta = _utcnow() - started
+            elapsed += int(delta.total_seconds())
+        return elapsed
+
+    def _remaining(self, session: PomodoroSession) -> int:
+        total = session.duration_min * 60
+        return max(0, total - self._elapsed(session))
+
+    def _try_auto_complete(self, session: PomodoroSession) -> bool:
+        if session.status != "running":
+            return False
+        if self._remaining(session) > 0:
+            return False
+        self._finalize_session(session, auto=True)
+        return True
+
+    def _finalize_session(
+        self,
+        session: PomodoroSession,
+        *,
+        auto: bool,
+        result: str = "",
+        completed: bool | None = None,
+        cancelled: bool = False,
+    ) -> None:
+        session.elapsed_seconds = self._elapsed(session)
+        session.started_at = None
+        session.finished_at = _utcnow()
+        session.completion_notified = False
+        session.result = result or None
+
+        if cancelled:
+            session.status = "cancelled"
+            session.completed = False
+        elif completed is not None:
+            session.status = "completed"
+            session.completed = completed
+        else:
+            session.status = "completed"
+            session.completed = True
+
+        self.db.commit()
+        self.db.refresh(session)
+
+    def _start_phase(
+        self,
+        phase: str,
+        *,
+        duration_min: int | None = None,
+        task_note: str | None = None,
+    ) -> PomodoroSession:
+        active = self._get_active()
+        if active:
+            raise ValueError("Таймер уже запущен. Сначала остановите текущую сессию.")
+
+        cycle = self.cycle.get()
+        if task_note is not None:
+            cycle.task_note = task_note
+        elif phase == PHASE_WORK and not cycle.task_note:
+            cycle.task_note = ""
+
+        duration = duration_min or self.cycle.duration_for_phase(phase, cycle)
+        note = task_note if task_note is not None else cycle.task_note
+
+        session = PomodoroSession(
+            user_id=self.user_id,
+            status="running",
+            phase=phase,
+            duration_min=duration,
+            task_note=note,
+            started_at=_utcnow(),
+        )
+        self.db.add(session)
+        self.db.commit()
+        self.db.refresh(session)
+        return session
+
+    def _to_status_dict(self, session: PomodoroSession | None) -> dict:
+        cycle_dict = self.cycle.to_dict()
+        if not session:
+            return {
+                "status": "idle",
+                "phase": PHASE_WORK,
+                "duration_min": cycle_dict["work_duration_min"],
+                "task_note": cycle_dict["task_note"],
+                "elapsed_seconds": 0,
+                "remaining_seconds": 0,
+                "session_id": None,
+                "cycle": cycle_dict,
+            }
+
+        elapsed = self._elapsed(session)
+        total = session.duration_min * 60
+        remaining = max(0, total - elapsed)
+
+        return {
+            "status": session.status,
+            "phase": session.phase,
+            "duration_min": session.duration_min,
+            "task_note": session.task_note,
+            "elapsed_seconds": elapsed,
+            "remaining_seconds": remaining,
+            "session_id": session.id,
+            "started_at": session.started_at.isoformat() if session.started_at else None,
+            "finished_at": session.finished_at.isoformat() if session.finished_at else None,
+            "cycle": cycle_dict,
+        }
+
+    def get_status(self) -> dict:
+        active = self._get_active()
+        if active:
+            self._try_auto_complete(active)
+            active = self._get_active()
+        return self._to_status_dict(active)
+
+    def start_work(self, duration_min: int | None = None, task_note: str = "") -> dict:
+        session = self._start_phase(
+            PHASE_WORK,
+            duration_min=duration_min,
+            task_note=task_note,
+        )
+        return self._to_status_dict(session)
+
+    def start_short_break(self, duration_min: int | None = None) -> dict:
+        session = self._start_phase(PHASE_SHORT_BREAK, duration_min=duration_min)
+        return self._to_status_dict(session)
+
+    def start_long_break(self, duration_min: int | None = None) -> dict:
+        session = self._start_phase(PHASE_LONG_BREAK, duration_min=duration_min)
+        return self._to_status_dict(session)
+
+    def start(self, duration_min: int = 25, task_note: str = "") -> dict:
+        return self.start_work(duration_min=duration_min, task_note=task_note)
+
+    def pause(self) -> dict:
+        session = self._get_active()
+        if not session or session.status != "running":
+            raise ValueError("Нет активного запущенного таймера.")
+
+        session.elapsed_seconds = self._elapsed(session)
+        session.status = "paused"
+        session.paused_at = _utcnow()
+        session.started_at = None
+        self.db.commit()
+        self.db.refresh(session)
+        return self._to_status_dict(session)
+
+    def resume(self) -> dict:
+        session = self._get_active()
+        if not session or session.status != "paused":
+            raise ValueError("Нет таймера на паузе.")
+
+        session.status = "running"
+        session.started_at = _utcnow()
+        session.paused_at = None
+        self.db.commit()
+        self.db.refresh(session)
+        return self._to_status_dict(session)
+
+    def stop(self, result: str = "", completed: bool = False) -> dict:
+        session = self._get_active()
+        if not session:
+            raise ValueError("Нет активного таймера.")
+
+        if completed:
+            self._finalize_session(session, auto=False, result=result, completed=True)
+        else:
+            self._finalize_session(session, auto=False, result=result, cancelled=True)
+            session.completion_notified = True
+            self.db.commit()
+        return self._to_status_dict(None)
+
+    def reset_cycle(self, clear_task: bool = False) -> dict:
+        active = self._get_active()
+        if active:
+            self._finalize_session(active, auto=False, cancelled=True)
+            active.completion_notified = True
+            self.db.commit()
+        cycle = self.cycle.reset(clear_task=clear_task)
+        status = self._to_status_dict(None)
+        status["cycle"] = cycle
+        return status
+
+    def skip_phase(self) -> dict:
+        session = self._get_active()
+        if not session:
+            raise ValueError("Нет активного таймера.")
+
+        self._finalize_session(session, auto=True)
+        return self._to_status_dict(None)
+
+    def get_pending_completions(self) -> list[PomodoroSession]:
+        stmt = (
+            select(PomodoroSession)
+            .where(
+                PomodoroSession.user_id == self.user_id,
+                PomodoroSession.status == "completed",
+                PomodoroSession.completed.is_(True),
+                PomodoroSession.completion_notified.is_(False),
+            )
+            .order_by(PomodoroSession.id.asc())
+        )
+        return list(self.db.scalars(stmt))
+
+    def mark_notified(self, session: PomodoroSession) -> None:
+        session.completion_notified = True
+        self.db.commit()
+
+    def advance_after_completion(self, session: PomodoroSession) -> dict | None:
+        """Update cycle counters and auto-start next phase. Returns new status or None."""
+        phase = session.phase
+        cycle = self.cycle.get()
+
+        if phase == PHASE_WORK:
+            next_phase = self.cycle.on_work_completed()
+        elif phase == PHASE_SHORT_BREAK:
+            next_phase = PHASE_WORK
+        elif phase == PHASE_LONG_BREAK:
+            self.cycle.on_long_break_completed()
+            next_phase = None
+        else:
+            next_phase = None
+
+        if not cycle.auto_advance or next_phase is None:
+            return None
+
+        new_session = self._start_phase(next_phase)
+        return self._to_status_dict(new_session)
+
+    def history(self, limit: int = 20) -> list[dict]:
+        stmt = (
+            select(PomodoroSession)
+            .where(
+                PomodoroSession.user_id == self.user_id,
+                PomodoroSession.status.in_(("completed", "cancelled")),
+            )
+            .order_by(PomodoroSession.finished_at.desc())
+            .limit(limit)
+        )
+        sessions = self.db.scalars(stmt).all()
+        return [
+            {
+                "id": s.id,
+                "status": s.status,
+                "phase": s.phase,
+                "duration_min": s.duration_min,
+                "task_note": s.task_note,
+                "result": s.result,
+                "completed": s.completed,
+                "elapsed_seconds": s.elapsed_seconds,
+                "finished_at": s.finished_at.isoformat() if s.finished_at else None,
+            }
+            for s in sessions
+        ]

backend/app/pomodoro/watcher.py

@@ -1,38 +1,41 @@
-import asyncio
-import logging
-
-from app.db.base import SessionLocal
-from app.pomodoro.completion import PomodoroCompletionHandler
-from app.pomodoro.service import PomodoroService
-
-logger = logging.getLogger(__name__)
-
-WATCH_INTERVAL_SEC = 2
-
-
-async def pomodoro_watcher_loop() -> None:
-    while True:
-        try:
-            await asyncio.sleep(WATCH_INTERVAL_SEC)
-            await _tick()
-        except asyncio.CancelledError:
-            raise
-        except Exception:
-            logger.exception("Pomodoro watcher error")
-
-
-async def _tick() -> None:
-    db = SessionLocal()
-    try:
-        service = PomodoroService(db)
-        service.get_status()
-
-        pending = service.get_pending_completions()
-        if not pending:
-            return
-
-        handler = PomodoroCompletionHandler(db)
-        for session in pending:
-            await handler.process(session)
-    finally:
-        db.close()
+import asyncio
+import logging
+
+from sqlalchemy import select
+
+from app.db.base import SessionLocal
+from app.db.models import User
+from app.pomodoro.completion import PomodoroCompletionHandler
+from app.pomodoro.service import PomodoroService
+
+logger = logging.getLogger(__name__)
+
+WATCH_INTERVAL_SEC = 2
+
+
+async def pomodoro_watcher_loop() -> None:
+    while True:
+        try:
+            await asyncio.sleep(WATCH_INTERVAL_SEC)
+            await _tick()
+        except asyncio.CancelledError:
+            raise
+        except Exception:
+            logger.exception("Pomodoro watcher error")
+
+
+async def _tick() -> None:
+    db = SessionLocal()
+    try:
+        users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+        for user in users:
+            service = PomodoroService(db, user.id)
+            service.get_status()
+            pending = service.get_pending_completions()
+            if not pending:
+                continue
+            handler = PomodoroCompletionHandler(db, user.id)
+            for session in pending:
+                await handler.process(session)
+    finally:
+        db.close()

backend/app/projects/context.py

@@ -1,153 +1,155 @@
-import time
-from typing import Any
-
-from sqlalchemy.orm import Session
-
-from app.config import get_settings
-from app.integrations.taiga import TaigaClient
-from app.projects.service import ProjectService
-
-MAX_PROJECTS_IN_CONTEXT = 20
-MAX_OPEN_PER_PROJECT = 8
-PROJECTS_CACHE_SEC = 120
-
-_cache: dict[str, Any] = {"data": None, "expires_at": 0.0}
-
-
-def invalidate_projects_snapshot_cache() -> None:
-    _cache["data"] = None
-    _cache["expires_at"] = 0.0
-
-
-def get_projects_snapshot(db: Session, *, force: bool = False) -> dict[str, Any]:
-    now = time.time()
-    if not force and _cache["data"] is not None and now < _cache["expires_at"]:
-        return _cache["data"]
-
-    snapshot = _fetch_projects_snapshot(db)
-    _cache["data"] = snapshot
-    _cache["expires_at"] = now + PROJECTS_CACHE_SEC
-    return snapshot
-
-
-def _fetch_projects_snapshot(db: Session) -> dict[str, Any]:
-    settings = get_settings()
-    service = ProjectService(db)
-
-    if not settings.taiga_configured:
-        return {"configured": False, "projects": [], "open_items": [], "taiga_open": []}
-
-    projects = service.list_projects()
-    if not projects:
-        try:
-            projects = service.sync_taiga_projects()
-        except Exception as exc:
-            return {
-                "configured": True,
-                "projects": [],
-                "open_items": [],
-                "taiga_open": [],
-                "error": str(exc),
-            }
-
-    open_items = service.list_work_items(limit=15, status="open")
-    taiga_open: list[dict[str, Any]] = []
-    fetch_error: str | None = None
-
-    try:
-        client = TaigaClient()
-        for proj in projects[:MAX_PROJECTS_IN_CONTEXT]:
-            stories = client.list_open_userstories(
-                proj["taiga_id"], limit=MAX_OPEN_PER_PROJECT
-            )
-            tasks = client.list_open_tasks(proj["taiga_id"], limit=MAX_OPEN_PER_PROJECT)
-            taiga_open.append(
-                {
-                    "slug": proj["slug"],
-                    "name": proj["name"],
-                    "stories": [
-                        {
-                            "ref": s.get("ref"),
-                            "subject": s.get("subject", "")[:120],
-                        }
-                        for s in stories
-                    ],
-                    "tasks": [
-                        {
-                            "ref": t.get("ref"),
-                            "subject": t.get("subject", "")[:120],
-                        }
-                        for t in tasks
-                    ],
-                }
-            )
-    except Exception as exc:
-        fetch_error = str(exc)
-
-    return {
-        "configured": True,
-        "projects": projects,
-        "open_items": open_items,
-        "taiga_open": taiga_open,
-        "error": fetch_error,
-    }
-
-
-def format_projects_context(snapshot: dict[str, Any]) -> str:
-    if not snapshot.get("configured"):
-        return "[Taiga/Gitea]\nНе настроено (нет TAIGA_USERNAME/PASSWORD в .env)."
-
-    lines = ["[Проекты и задачи — снимок на начало ответа]"]
-
-    if snapshot.get("error"):
-        lines.append(f"⚠ Ошибка загрузки задач из Taiga: {snapshot['error']}")
-
-    projects = snapshot.get("projects") or []
-    if not projects:
-        lines.append("Проекты Taiga: кэш пуст. Вызови sync_taiga_projects.")
-    else:
-        lines.append(f"Проекты Taiga ({len(projects)}):")
-        for p in projects[:MAX_PROJECTS_IN_CONTEXT]:
-            gitea = (
-                f"{p.get('gitea_owner')}/{p.get('gitea_repo')}"
-                if p.get("gitea_configured")
-                else "Gitea не привязан"
-            )
-            lines.append(f"- `{p.get('slug')}`: {p.get('name')} · {gitea}")
-
-    taiga_open = snapshot.get("taiga_open") or []
-    if taiga_open:
-        lines.append("")
-        lines.append("Открытые задачи в Taiga (live):")
-        for block in taiga_open:
-            stories = block.get("stories") or []
-            tasks = block.get("tasks") or []
-            if not stories and not tasks:
-                lines.append(f"  `{block.get('slug')}`: нет открытых")
-                continue
-            lines.append(f"  `{block.get('slug')}`:")
-            for story in stories:
-                lines.append(f"    story #{story.get('ref')} {story.get('subject')}")
-            for task in tasks:
-                lines.append(f"    task #{task.get('ref')} {task.get('subject')}")
-
-    open_items = snapshot.get("open_items") or []
-    if open_items:
-        lines.append("")
-        lines.append("Work items созданные ассистентом (локальная БД):")
-        for item in open_items[:10]:
-            gitea_part = f", gitea #{item.get('gitea_issue')}" if item.get("gitea_issue") else ""
-            lines.append(
-                f"- #{item.get('taiga_ref')} {item.get('title')} "
-                f"({item.get('taiga_slug')}{gitea_part})"
-            )
-
-    lines.append("")
-    lines.append(
-        "Правила: "
-        "«какие задачи» → list_taiga_tasks (Taiga API), НЕ list_work_items. "
-        "list_work_items — только созданные через ассистента. "
-        "Не пиши «ожидаю систему» — сразу вызывай tool или отвечай из снимка выше. "
-        "create_work_item — для новых фич/багов из вольного текста."
-    )
-    return "\n".join(lines)
+import time
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.config import get_settings
+from app.integrations.taiga import TaigaClient
+from app.projects.service import ProjectService
+
+MAX_PROJECTS_IN_CONTEXT = 20
+MAX_OPEN_PER_PROJECT = 8
+PROJECTS_CACHE_SEC = 120
+
+_cache: dict[int, dict[str, Any]] = {}
+
+
+def invalidate_projects_snapshot_cache(user_id: int | None = None) -> None:
+    if user_id is None:
+        _cache.clear()
+    else:
+        _cache.pop(user_id, None)
+
+
+def get_projects_snapshot(db: Session, user_id: int, *, force: bool = False) -> dict[str, Any]:
+    now = time.time()
+    entry = _cache.get(user_id)
+    if not force and entry and now < entry.get("expires_at", 0):
+        return entry["data"]
+
+    snapshot = _fetch_projects_snapshot(db, user_id)
+    _cache[user_id] = {"data": snapshot, "expires_at": now + PROJECTS_CACHE_SEC}
+    return snapshot
+
+
+def _fetch_projects_snapshot(db: Session, user_id: int) -> dict[str, Any]:
+    settings = get_settings()
+    service = ProjectService(db, user_id)
+
+    if not settings.taiga_configured:
+        return {"configured": False, "projects": [], "open_items": [], "taiga_open": []}
+
+    projects = service.list_projects()
+    if not projects:
+        try:
+            projects = service.sync_taiga_projects()
+        except Exception as exc:
+            return {
+                "configured": True,
+                "projects": [],
+                "open_items": [],
+                "taiga_open": [],
+                "error": str(exc),
+            }
+
+    open_items = service.list_work_items(limit=15, status="open")
+    taiga_open: list[dict[str, Any]] = []
+    fetch_error: str | None = None
+
+    try:
+        client = TaigaClient()
+        for proj in projects[:MAX_PROJECTS_IN_CONTEXT]:
+            stories = client.list_open_userstories(
+                proj["taiga_id"], limit=MAX_OPEN_PER_PROJECT
+            )
+            tasks = client.list_open_tasks(proj["taiga_id"], limit=MAX_OPEN_PER_PROJECT)
+            taiga_open.append(
+                {
+                    "slug": proj["slug"],
+                    "name": proj["name"],
+                    "stories": [
+                        {
+                            "ref": s.get("ref"),
+                            "subject": s.get("subject", "")[:120],
+                        }
+                        for s in stories
+                    ],
+                    "tasks": [
+                        {
+                            "ref": t.get("ref"),
+                            "subject": t.get("subject", "")[:120],
+                        }
+                        for t in tasks
+                    ],
+                }
+            )
+    except Exception as exc:
+        fetch_error = str(exc)
+
+    return {
+        "configured": True,
+        "projects": projects,
+        "open_items": open_items,
+        "taiga_open": taiga_open,
+        "error": fetch_error,
+    }
+
+
+def format_projects_context(snapshot: dict[str, Any]) -> str:
+    if not snapshot.get("configured"):
+        return "[Taiga/Gitea]\nНе настроено (нет TAIGA_USERNAME/PASSWORD в .env)."
+
+    lines = ["[Проекты и задачи — снимок на начало ответа]"]
+
+    if snapshot.get("error"):
+        lines.append(f"⚠ Ошибка загрузки задач из Taiga: {snapshot['error']}")
+
+    projects = snapshot.get("projects") or []
+    if not projects:
+        lines.append("Проекты Taiga: кэш пуст. Вызови sync_taiga_projects.")
+    else:
+        lines.append(f"Проекты Taiga ({len(projects)}):")
+        for p in projects[:MAX_PROJECTS_IN_CONTEXT]:
+            gitea = (
+                f"{p.get('gitea_owner')}/{p.get('gitea_repo')}"
+                if p.get("gitea_configured")
+                else "Gitea не привязан"
+            )
+            lines.append(f"- `{p.get('slug')}`: {p.get('name')} · {gitea}")
+
+    taiga_open = snapshot.get("taiga_open") or []
+    if taiga_open:
+        lines.append("")
+        lines.append("Открытые задачи в Taiga (live):")
+        for block in taiga_open:
+            stories = block.get("stories") or []
+            tasks = block.get("tasks") or []
+            if not stories and not tasks:
+                lines.append(f"  `{block.get('slug')}`: нет открытых")
+                continue
+            lines.append(f"  `{block.get('slug')}`:")
+            for story in stories:
+                lines.append(f"    story #{story.get('ref')} {story.get('subject')}")
+            for task in tasks:
+                lines.append(f"    task #{task.get('ref')} {task.get('subject')}")
+
+    open_items = snapshot.get("open_items") or []
+    if open_items:
+        lines.append("")
+        lines.append("Work items созданные ассистентом (локальная БД):")
+        for item in open_items[:10]:
+            gitea_part = f", gitea #{item.get('gitea_issue')}" if item.get("gitea_issue") else ""
+            lines.append(
+                f"- #{item.get('taiga_ref')} {item.get('title')} "
+                f"({item.get('taiga_slug')}{gitea_part})"
+            )
+
+    lines.append("")
+    lines.append(
+        "Правила: "
+        "«какие задачи» → list_taiga_tasks (Taiga API), НЕ list_work_items. "
+        "list_work_items — только созданные через ассистента. "
+        "Не пиши «ожидаю систему» — сразу вызывай tool или отвечай из снимка выше. "
+        "create_work_item — для новых фич/багов из вольного текста."
+    )
+    return "\n".join(lines)

backend/app/projects/service.py

@@ -1,466 +1,476 @@
-from datetime import datetime, timezone
-from typing import Any
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.config import get_settings
-from app.db.models import ProjectBinding, TaigaProject, WorkItem
-from app.integrations.gitea import GiteaClient
-from app.integrations.taiga import TaigaClient
-from app.projects.commit_parser import parse_commit_message
-from app.projects.structuring import (
-    format_gitea_body,
-    format_story_description,
-    slugify_branch,
-    structure_work_item,
-)
-
-
-class ProjectService:
-    def __init__(self, db: Session):
-        self.db = db
-        self.settings = get_settings()
-
-    def sync_taiga_projects(self) -> list[dict[str, Any]]:
-        if not self.settings.taiga_configured:
-            raise ValueError("Taiga не настроена: задайте TAIGA_USERNAME и TAIGA_PASSWORD")
-
-        client = TaigaClient()
-        remote = client.list_projects()
-        now = datetime.now(timezone.utc)
-
-        for item in remote:
-            slug = item.get("slug") or ""
-            if not slug:
-                continue
-            existing = self.db.scalar(
-                select(TaigaProject).where(TaigaProject.slug == slug)
-            )
-            if existing:
-                existing.name = item.get("name", slug)
-                existing.taiga_id = item["id"]
-                existing.synced_at = now
-            else:
-                self.db.add(
-                    TaigaProject(
-                        taiga_id=item["id"],
-                        name=item.get("name", slug),
-                        slug=slug,
-                        synced_at=now,
-                    )
-                )
-        self.db.commit()
-        return self.list_projects()
-
-    def list_projects(self) -> list[dict[str, Any]]:
-        stmt = (
-            select(TaigaProject, ProjectBinding)
-            .outerjoin(ProjectBinding, ProjectBinding.taiga_slug == TaigaProject.slug)
-            .order_by(TaigaProject.name)
-        )
-        rows = self.db.execute(stmt).all()
-        result = []
-        for taiga_proj, binding in rows:
-            result.append(
-                {
-                    "taiga_id": taiga_proj.taiga_id,
-                    "name": taiga_proj.name,
-                    "slug": taiga_proj.slug,
-                    "gitea_owner": binding.gitea_owner if binding else "",
-                    "gitea_repo": binding.gitea_repo if binding else "",
-                    "default_branch": binding.default_branch if binding else "main",
-                    "gitea_configured": bool(binding and binding.gitea_owner and binding.gitea_repo),
-                }
-            )
-        return result
-
-    def bind_gitea(
-        self,
-        taiga_slug: str,
-        gitea_owner: str,
-        gitea_repo: str,
-        default_branch: str = "main",
-    ) -> dict[str, Any]:
-        if not self.db.scalar(select(TaigaProject).where(TaigaProject.slug == taiga_slug)):
-            raise ValueError(f"Проект Taiga '{taiga_slug}' не найден. Сначала sync-taiga.")
-
-        binding = self.db.scalar(
-            select(ProjectBinding).where(ProjectBinding.taiga_slug == taiga_slug)
-        )
-        if binding:
-            binding.gitea_owner = gitea_owner
-            binding.gitea_repo = gitea_repo
-            binding.default_branch = default_branch
-        else:
-            binding = ProjectBinding(
-                taiga_slug=taiga_slug,
-                gitea_owner=gitea_owner,
-                gitea_repo=gitea_repo,
-                default_branch=default_branch,
-            )
-            self.db.add(binding)
-        self.db.commit()
-
-        for proj in self.list_projects():
-            if proj["slug"] == taiga_slug:
-                return proj
-        raise ValueError("Binding failed")
-
-    def _resolve_project(self, slug: str | None) -> tuple[TaigaProject, ProjectBinding | None]:
-        projects = self.db.scalars(select(TaigaProject).order_by(TaigaProject.name)).all()
-        if not projects:
-            raise ValueError("Нет проектов Taiga. Вызовите sync_taiga_projects.")
-
-        taiga_proj: TaigaProject | None = None
-        if slug:
-            taiga_proj = self.db.scalar(
-                select(TaigaProject).where(TaigaProject.slug == slug)
-            )
-            if not taiga_proj:
-                raise ValueError(f"Проект '{slug}' не найден")
-        else:
-            taiga_proj = projects[0]
-
-        binding = self.db.scalar(
-            select(ProjectBinding).where(ProjectBinding.taiga_slug == taiga_proj.slug)
-        )
-        return taiga_proj, binding
-
-    async def create_work_item(
-        self, raw_text: str, project_slug: str | None = None
-    ) -> dict[str, Any]:
-        if not self.settings.taiga_configured:
-            raise ValueError("Taiga не настроена")
-
-        project_list = self.list_projects()
-        if not project_list:
-            self.sync_taiga_projects()
-            project_list = self.list_projects()
-
-        structured = await structure_work_item(raw_text, project_list)
-        slug = project_slug or structured.get("project_slug")
-        taiga_proj, binding = self._resolve_project(slug)
-
-        if binding and not (binding.gitea_owner and binding.gitea_repo):
-            binding = None
-
-        taiga = TaigaClient()
-        title = (structured.get("title") or raw_text).strip()[:500]
-        description = format_story_description(structured, raw_text)
-        tags = structured.get("tags") or []
-        issue_type = structured.get("issue_type", "feature")
-        if issue_type == "bug" and "bug" not in [t.lower() for t in tags]:
-            tags.append("bug")
-
-        story = taiga.create_userstory(
-            taiga_proj.taiga_id,
-            title,
-            description,
-            tags=tags,
-        )
-
-        subtasks = []
-        for child in structured.get("children") or []:
-            if isinstance(child, dict):
-                subtasks.append(
-                    taiga.create_task(
-                        taiga_proj.taiga_id,
-                        story["id"],
-                        child.get("title", "Подзадача"),
-                        child.get("description", ""),
-                    )
-                )
-
-        branch = f"feature/{story['ref']}-{slugify_branch(title)}"
-        gitea_issue_number = None
-        gitea_url = ""
-
-        if binding and self.settings.gitea_configured:
-            gitea = GiteaClient()
-            gitea_body = format_gitea_body(
-                structured,
-                raw_text,
-                story["ref"],
-                taiga.story_url(taiga_proj.taiga_id, story["ref"]),
-                branch,
-            )
-            if issue_type:
-                gitea_body = f"**Тип:** {issue_type}\n\n{gitea_body}"
-            issue = gitea.create_issue(
-                binding.gitea_owner,
-                binding.gitea_repo,
-                title,
-                gitea_body,
-            )
-            gitea_issue_number = issue["number"]
-            gitea_url = gitea.issue_url(
-                binding.gitea_owner, binding.gitea_repo, gitea_issue_number
-            )
-
-        work_item = WorkItem(
-            taiga_slug=taiga_proj.slug,
-            taiga_project_id=taiga_proj.taiga_id,
-            taiga_story_id=story["id"],
-            taiga_story_ref=story["ref"],
-            gitea_owner=binding.gitea_owner if binding else "",
-            gitea_repo=binding.gitea_repo if binding else "",
-            gitea_issue_number=gitea_issue_number,
-            suggested_branch=branch,
-            raw_text=raw_text,
-            title=title,
-            status="open",
-        )
-        self.db.add(work_item)
-        self.db.commit()
-        self.db.refresh(work_item)
-
-        return {
-            "ok": True,
-            "work_item_id": work_item.id,
-            "taiga": {
-                "ref": story["ref"],
-                "id": story["id"],
-                "subject": story["subject"],
-                "url": taiga.story_url(taiga_proj.taiga_id, story["ref"]),
-            },
-            "gitea": {
-                "number": gitea_issue_number,
-                "url": gitea_url,
-            },
-            "branch": branch,
-            "issue_type": issue_type,
-            "subtasks": [{"ref": t.get("ref"), "subject": t.get("subject")} for t in subtasks],
-            "questions": structured.get("questions") or [],
-            "project_slug": taiga_proj.slug,
-        }
-
-    def process_push(
-        self, owner: str, repo: str, commits: list[dict[str, Any]]
-    ) -> list[dict[str, Any]]:
-        if not self.settings.taiga_configured:
-            return []
-
-        taiga = TaigaClient()
-        gitea = GiteaClient() if self.settings.gitea_configured else None
-        results: list[dict[str, Any]] = []
-
-        for commit in commits:
-            message = commit.get("message", "")
-            parsed = parse_commit_message(message)
-            sha = commit.get("id", "")[:8]
-
-            gitea_refs = set(parsed["gitea"])
-            taiga_story_refs = set(parsed["taiga_story"])
-            taiga_task_refs = set(parsed["taiga_task"])
-
-            linked_items = self.db.scalars(
-                select(WorkItem).where(
-                    WorkItem.gitea_owner == owner,
-                    WorkItem.gitea_repo == repo,
-                    WorkItem.status == "open",
-                )
-            ).all()
-
-            for item in linked_items:
-                if item.gitea_issue_number and item.gitea_issue_number in gitea_refs:
-                    taiga_story_refs.add(item.taiga_story_ref)
-                if item.taiga_story_ref in taiga_story_refs and item.gitea_issue_number:
-                    gitea_refs.add(item.gitea_issue_number)
-
-            for gitea_num in gitea_refs:
-                if gitea:
-                    try:
-                        gitea.close_issue(owner, repo, gitea_num)
-                    except Exception as exc:
-                        results.append({"error": f"gitea #{gitea_num}: {exc}"})
-                        continue
-
-                for item in linked_items:
-                    if item.gitea_issue_number == gitea_num:
-                        try:
-                            self._close_work_item(item, taiga)
-                            results.append(
-                                {
-                                    "commit": sha,
-                                    "closed": f"gitea #{gitea_num}, taiga #{item.taiga_story_ref}",
-                                }
-                            )
-                        except Exception as exc:
-                            results.append(
-                                {"error": f"work item {item.id} (gitea #{gitea_num}): {exc}"}
-                            )
-
-            for ref in taiga_story_refs:
-                project_id = self._project_id_for_ref(owner, repo, ref, linked_items)
-                if not project_id:
-                    continue
-                story = taiga.get_by_ref(project_id, ref, kind="userstory")
-                if story and not story.get("is_closed"):
-                    try:
-                        taiga.close_userstory(story["id"], project_id)
-                        results.append({"commit": sha, "closed": f"taiga #{ref}"})
-                    except Exception as exc:
-                        results.append({"error": f"taiga #{ref}: {exc}"})
-                for item in linked_items:
-                    if item.taiga_story_ref == ref and item.status != "closed":
-                        try:
-                            self._close_work_item(item, taiga, close_gitea=bool(gitea))
-                        except Exception as exc:
-                            results.append(
-                                {"error": f"work item {item.id} (taiga #{ref}): {exc}"}
-                            )
-
-            for ref in taiga_task_refs:
-                binding = self.db.scalar(
-                    select(ProjectBinding).where(
-                        ProjectBinding.gitea_owner == owner,
-                        ProjectBinding.gitea_repo == repo,
-                    )
-                )
-                if not binding:
-                    continue
-                taiga_proj = self.db.scalar(
-                    select(TaigaProject).where(TaigaProject.slug == binding.taiga_slug)
-                )
-                if not taiga_proj:
-                    continue
-                task = taiga.get_by_ref(taiga_proj.taiga_id, ref, kind="task")
-                if task and not task.get("is_closed"):
-                    try:
-                        taiga.close_task(task["id"], taiga_proj.taiga_id)
-                        results.append({"commit": sha, "closed": f"taiga task #{ref}"})
-                    except Exception as exc:
-                        results.append({"error": f"taiga task #{ref}: {exc}"})
-
-        self.db.commit()
-        return results
-
-    def _project_id_for_ref(
-        self,
-        owner: str,
-        repo: str,
-        ref: int,
-        items: list[WorkItem],
-    ) -> int | None:
-        for item in items:
-            if item.taiga_story_ref == ref:
-                return item.taiga_project_id
-        binding = self.db.scalar(
-            select(ProjectBinding).where(
-                ProjectBinding.gitea_owner == owner,
-                ProjectBinding.gitea_repo == repo,
-            )
-        )
-        if binding:
-            taiga_proj = self.db.scalar(
-                select(TaigaProject).where(TaigaProject.slug == binding.taiga_slug)
-            )
-            return taiga_proj.taiga_id if taiga_proj else None
-        return None
-
-    def _close_work_item(
-        self,
-        item: WorkItem,
-        taiga: TaigaClient,
-        *,
-        close_gitea: bool = True,
-    ) -> None:
-        if item.status == "closed":
-            return
-        story = taiga.get_by_ref(item.taiga_project_id, item.taiga_story_ref, kind="userstory")
-        if story:
-            taiga.close_userstory(story["id"], item.taiga_project_id)
-        if (
-            close_gitea
-            and item.gitea_issue_number
-            and self.settings.gitea_configured
-        ):
-            GiteaClient().close_issue(
-                item.gitea_owner, item.gitea_repo, item.gitea_issue_number
-            )
-        item.status = "closed"
-        item.closed_at = datetime.now(timezone.utc)
-
-    def list_taiga_open_tasks(
-        self,
-        project_slug: str | None = None,
-        limit: int = 20,
-    ) -> dict[str, Any]:
-        if not self.settings.taiga_configured:
-            raise ValueError("Taiga не настроена")
-
-        projects = self.list_projects()
-        if not projects:
-            projects = self.sync_taiga_projects()
-
-        if project_slug:
-            projects = [p for p in projects if p["slug"] == project_slug]
-            if not projects:
-                raise ValueError(
-                    f"Проект '{project_slug}' не найден. Вызови sync_taiga_projects."
-                )
-
-        client = TaigaClient()
-        blocks: list[dict[str, Any]] = []
-
-        for proj in projects:
-            stories = client.list_open_userstories(proj["taiga_id"], limit=limit)
-            tasks = client.list_open_tasks(proj["taiga_id"], limit=limit)
-            blocks.append(
-                {
-                    "slug": proj["slug"],
-                    "name": proj["name"],
-                    "taiga_id": proj["taiga_id"],
-                    "stories": [
-                        {
-                            "ref": s.get("ref"),
-                            "subject": s.get("subject", ""),
-                            "url": client.story_url(proj["taiga_id"], s.get("ref", 0)),
-                        }
-                        for s in stories
-                    ],
-                    "tasks": [
-                        {
-                            "ref": t.get("ref"),
-                            "subject": t.get("subject", ""),
-                            "user_story": t.get("user_story"),
-                        }
-                        for t in tasks
-                    ],
-                }
-            )
-
-        total_stories = sum(len(b["stories"]) for b in blocks)
-        total_tasks = sum(len(b["tasks"]) for b in blocks)
-        return {
-            "projects": blocks,
-            "total_stories": total_stories,
-            "total_tasks": total_tasks,
-        }
-
-    def list_work_items(self, limit: int = 30, status: str | None = None) -> list[dict[str, Any]]:
-        stmt = select(WorkItem).order_by(WorkItem.created_at.desc()).limit(limit)
-        if status:
-            stmt = stmt.where(WorkItem.status == status)
-        items = self.db.scalars(stmt).all()
-        settings = get_settings()
-        return [
-            {
-                "id": i.id,
-                "title": i.title,
-                "status": i.status,
-                "taiga_slug": i.taiga_slug,
-                "taiga_ref": i.taiga_story_ref,
-                "gitea_issue": i.gitea_issue_number,
-                "branch": i.suggested_branch,
-                "taiga_url": f"{settings.taiga_public_url}/project/0/{i.taiga_project_id}/us/{i.taiga_story_ref}",
-                "gitea_url": (
-                    f"{settings.gitea_public_url}/{i.gitea_owner}/{i.gitea_repo}/issues/{i.gitea_issue_number}"
-                    if i.gitea_issue_number
-                    else ""
-                ),
-                "created_at": i.created_at.isoformat() if i.created_at else None,
-            }
-            for i in items
-        ]
+from datetime import datetime, timezone
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.config import get_settings
+from app.db.models import ProjectBinding, TaigaProject, WorkItem
+from app.integrations.gitea import GiteaClient
+from app.integrations.taiga import TaigaClient
+from app.projects.commit_parser import parse_commit_message
+from app.projects.structuring import (
+    format_gitea_body,
+    format_story_description,
+    slugify_branch,
+    structure_work_item,
+)
+
+
+class ProjectService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+        self.settings = get_settings()
+
+    def sync_taiga_projects(self) -> list[dict[str, Any]]:
+        if not self.settings.taiga_configured:
+            raise ValueError("Taiga не настроена: задайте TAIGA_USERNAME и TAIGA_PASSWORD")
+
+        client = TaigaClient()
+        remote = client.list_projects()
+        now = datetime.now(timezone.utc)
+
+        for item in remote:
+            slug = item.get("slug") or ""
+            if not slug:
+                continue
+            existing = self.db.scalar(
+                select(TaigaProject).where(TaigaProject.slug == slug)
+            )
+            if existing:
+                existing.name = item.get("name", slug)
+                existing.taiga_id = item["id"]
+                existing.synced_at = now
+            else:
+                self.db.add(
+                    TaigaProject(
+                        taiga_id=item["id"],
+                        name=item.get("name", slug),
+                        slug=slug,
+                        synced_at=now,
+                    )
+                )
+        self.db.commit()
+        return self.list_projects()
+
+    def list_projects(self) -> list[dict[str, Any]]:
+        stmt = (
+            select(TaigaProject, ProjectBinding)
+            .outerjoin(
+                ProjectBinding,
+                (ProjectBinding.taiga_slug == TaigaProject.slug)
+                & (ProjectBinding.user_id == self.user_id),
+            )
+            .order_by(TaigaProject.name)
+        )
+        rows = self.db.execute(stmt).all()
+        result = []
+        for taiga_proj, binding in rows:
+            result.append(
+                {
+                    "taiga_id": taiga_proj.taiga_id,
+                    "name": taiga_proj.name,
+                    "slug": taiga_proj.slug,
+                    "gitea_owner": binding.gitea_owner if binding else "",
+                    "gitea_repo": binding.gitea_repo if binding else "",
+                    "default_branch": binding.default_branch if binding else "main",
+                    "gitea_configured": bool(binding and binding.gitea_owner and binding.gitea_repo),
+                }
+            )
+        return result
+
+    def bind_gitea(
+        self,
+        taiga_slug: str,
+        gitea_owner: str,
+        gitea_repo: str,
+        default_branch: str = "main",
+    ) -> dict[str, Any]:
+        if not self.db.scalar(select(TaigaProject).where(TaigaProject.slug == taiga_slug)):
+            raise ValueError(f"Проект Taiga '{taiga_slug}' не найден. Сначала sync-taiga.")
+
+        binding = self.db.scalar(
+            select(ProjectBinding).where(ProjectBinding.user_id == self.user_id, ProjectBinding.taiga_slug == taiga_slug)
+        )
+        if binding:
+            binding.gitea_owner = gitea_owner
+            binding.gitea_repo = gitea_repo
+            binding.default_branch = default_branch
+        else:
+            binding = ProjectBinding(
+                user_id=self.user_id,
+                taiga_slug=taiga_slug,
+                gitea_owner=gitea_owner,
+                gitea_repo=gitea_repo,
+                default_branch=default_branch,
+            )
+            self.db.add(binding)
+        self.db.commit()
+
+        for proj in self.list_projects():
+            if proj["slug"] == taiga_slug:
+                return proj
+        raise ValueError("Binding failed")
+
+    def _resolve_project(self, slug: str | None) -> tuple[TaigaProject, ProjectBinding | None]:
+        projects = self.db.scalars(select(TaigaProject).order_by(TaigaProject.name)).all()
+        if not projects:
+            raise ValueError("Нет проектов Taiga. Вызовите sync_taiga_projects.")
+
+        taiga_proj: TaigaProject | None = None
+        if slug:
+            taiga_proj = self.db.scalar(
+                select(TaigaProject).where(TaigaProject.slug == slug)
+            )
+            if not taiga_proj:
+                raise ValueError(f"Проект '{slug}' не найден")
+        else:
+            taiga_proj = projects[0]
+
+        binding = self.db.scalar(
+            select(ProjectBinding).where(ProjectBinding.user_id == self.user_id, ProjectBinding.taiga_slug == taiga_proj.slug)
+        )
+        return taiga_proj, binding
+
+    async def create_work_item(
+        self, raw_text: str, project_slug: str | None = None
+    ) -> dict[str, Any]:
+        if not self.settings.taiga_configured:
+            raise ValueError("Taiga не настроена")
+
+        project_list = self.list_projects()
+        if not project_list:
+            self.sync_taiga_projects()
+            project_list = self.list_projects()
+
+        structured = await structure_work_item(raw_text, project_list)
+        slug = project_slug or structured.get("project_slug")
+        taiga_proj, binding = self._resolve_project(slug)
+
+        if binding and not (binding.gitea_owner and binding.gitea_repo):
+            binding = None
+
+        taiga = TaigaClient()
+        title = (structured.get("title") or raw_text).strip()[:500]
+        description = format_story_description(structured, raw_text)
+        tags = structured.get("tags") or []
+        issue_type = structured.get("issue_type", "feature")
+        if issue_type == "bug" and "bug" not in [t.lower() for t in tags]:
+            tags.append("bug")
+
+        story = taiga.create_userstory(
+            taiga_proj.taiga_id,
+            title,
+            description,
+            tags=tags,
+        )
+
+        subtasks = []
+        for child in structured.get("children") or []:
+            if isinstance(child, dict):
+                subtasks.append(
+                    taiga.create_task(
+                        taiga_proj.taiga_id,
+                        story["id"],
+                        child.get("title", "Подзадача"),
+                        child.get("description", ""),
+                    )
+                )
+
+        branch = f"feature/{story['ref']}-{slugify_branch(title)}"
+        gitea_issue_number = None
+        gitea_url = ""
+
+        if binding and self.settings.gitea_configured:
+            gitea = GiteaClient()
+            gitea_body = format_gitea_body(
+                structured,
+                raw_text,
+                story["ref"],
+                taiga.story_url(taiga_proj.taiga_id, story["ref"]),
+                branch,
+            )
+            if issue_type:
+                gitea_body = f"**Тип:** {issue_type}\n\n{gitea_body}"
+            issue = gitea.create_issue(
+                binding.gitea_owner,
+                binding.gitea_repo,
+                title,
+                gitea_body,
+            )
+            gitea_issue_number = issue["number"]
+            gitea_url = gitea.issue_url(
+                binding.gitea_owner, binding.gitea_repo, gitea_issue_number
+            )
+
+        work_item = WorkItem(
+            user_id=self.user_id,
+            taiga_slug=taiga_proj.slug,
+            taiga_project_id=taiga_proj.taiga_id,
+            taiga_story_id=story["id"],
+            taiga_story_ref=story["ref"],
+            gitea_owner=binding.gitea_owner if binding else "",
+            gitea_repo=binding.gitea_repo if binding else "",
+            gitea_issue_number=gitea_issue_number,
+            suggested_branch=branch,
+            raw_text=raw_text,
+            title=title,
+            status="open",
+        )
+        self.db.add(work_item)
+        self.db.commit()
+        self.db.refresh(work_item)
+
+        return {
+            "ok": True,
+            "work_item_id": work_item.id,
+            "taiga": {
+                "ref": story["ref"],
+                "id": story["id"],
+                "subject": story["subject"],
+                "url": taiga.story_url(taiga_proj.taiga_id, story["ref"]),
+            },
+            "gitea": {
+                "number": gitea_issue_number,
+                "url": gitea_url,
+            },
+            "branch": branch,
+            "issue_type": issue_type,
+            "subtasks": [{"ref": t.get("ref"), "subject": t.get("subject")} for t in subtasks],
+            "questions": structured.get("questions") or [],
+            "project_slug": taiga_proj.slug,
+        }
+
+    def process_push(
+        self, owner: str, repo: str, commits: list[dict[str, Any]]
+    ) -> list[dict[str, Any]]:
+        if not self.settings.taiga_configured:
+            return []
+
+        taiga = TaigaClient()
+        gitea = GiteaClient() if self.settings.gitea_configured else None
+        results: list[dict[str, Any]] = []
+
+        for commit in commits:
+            message = commit.get("message", "")
+            parsed = parse_commit_message(message)
+            sha = commit.get("id", "")[:8]
+
+            gitea_refs = set(parsed["gitea"])
+            taiga_story_refs = set(parsed["taiga_story"])
+            taiga_task_refs = set(parsed["taiga_task"])
+
+            linked_items = self.db.scalars(
+                select(WorkItem).where(
+                    WorkItem.user_id == self.user_id,
+                    WorkItem.gitea_owner == owner,
+                    WorkItem.gitea_repo == repo,
+                    WorkItem.status == "open",
+                )
+            ).all()
+
+            for item in linked_items:
+                if item.gitea_issue_number and item.gitea_issue_number in gitea_refs:
+                    taiga_story_refs.add(item.taiga_story_ref)
+                if item.taiga_story_ref in taiga_story_refs and item.gitea_issue_number:
+                    gitea_refs.add(item.gitea_issue_number)
+
+            for gitea_num in gitea_refs:
+                if gitea:
+                    try:
+                        gitea.close_issue(owner, repo, gitea_num)
+                    except Exception as exc:
+                        results.append({"error": f"gitea #{gitea_num}: {exc}"})
+                        continue
+
+                for item in linked_items:
+                    if item.gitea_issue_number == gitea_num:
+                        try:
+                            self._close_work_item(item, taiga)
+                            results.append(
+                                {
+                                    "commit": sha,
+                                    "closed": f"gitea #{gitea_num}, taiga #{item.taiga_story_ref}",
+                                }
+                            )
+                        except Exception as exc:
+                            results.append(
+                                {"error": f"work item {item.id} (gitea #{gitea_num}): {exc}"}
+                            )
+
+            for ref in taiga_story_refs:
+                project_id = self._project_id_for_ref(owner, repo, ref, linked_items)
+                if not project_id:
+                    continue
+                story = taiga.get_by_ref(project_id, ref, kind="userstory")
+                if story and not story.get("is_closed"):
+                    try:
+                        taiga.close_userstory(story["id"], project_id)
+                        results.append({"commit": sha, "closed": f"taiga #{ref}"})
+                    except Exception as exc:
+                        results.append({"error": f"taiga #{ref}: {exc}"})
+                for item in linked_items:
+                    if item.taiga_story_ref == ref and item.status != "closed":
+                        try:
+                            self._close_work_item(item, taiga, close_gitea=bool(gitea))
+                        except Exception as exc:
+                            results.append(
+                                {"error": f"work item {item.id} (taiga #{ref}): {exc}"}
+                            )
+
+            for ref in taiga_task_refs:
+                binding = self.db.scalar(
+                    select(ProjectBinding).where(
+                        ProjectBinding.user_id == self.user_id,
+                        ProjectBinding.gitea_owner == owner,
+                        ProjectBinding.gitea_repo == repo,
+                    )
+                )
+                if not binding:
+                    continue
+                taiga_proj = self.db.scalar(
+                    select(TaigaProject).where(TaigaProject.slug == binding.taiga_slug)
+                )
+                if not taiga_proj:
+                    continue
+                task = taiga.get_by_ref(taiga_proj.taiga_id, ref, kind="task")
+                if task and not task.get("is_closed"):
+                    try:
+                        taiga.close_task(task["id"], taiga_proj.taiga_id)
+                        results.append({"commit": sha, "closed": f"taiga task #{ref}"})
+                    except Exception as exc:
+                        results.append({"error": f"taiga task #{ref}: {exc}"})
+
+        self.db.commit()
+        return results
+
+    def _project_id_for_ref(
+        self,
+        owner: str,
+        repo: str,
+        ref: int,
+        items: list[WorkItem],
+    ) -> int | None:
+        for item in items:
+            if item.taiga_story_ref == ref:
+                return item.taiga_project_id
+        binding = self.db.scalar(
+            select(ProjectBinding).where(
+                ProjectBinding.user_id == self.user_id,
+                ProjectBinding.gitea_owner == owner,
+                ProjectBinding.gitea_repo == repo,
+            )
+        )
+        if binding:
+            taiga_proj = self.db.scalar(
+                select(TaigaProject).where(TaigaProject.slug == binding.taiga_slug)
+            )
+            return taiga_proj.taiga_id if taiga_proj else None
+        return None
+
+    def _close_work_item(
+        self,
+        item: WorkItem,
+        taiga: TaigaClient,
+        *,
+        close_gitea: bool = True,
+    ) -> None:
+        if item.status == "closed":
+            return
+        story = taiga.get_by_ref(item.taiga_project_id, item.taiga_story_ref, kind="userstory")
+        if story:
+            taiga.close_userstory(story["id"], item.taiga_project_id)
+        if (
+            close_gitea
+            and item.gitea_issue_number
+            and self.settings.gitea_configured
+        ):
+            GiteaClient().close_issue(
+                item.gitea_owner, item.gitea_repo, item.gitea_issue_number
+            )
+        item.status = "closed"
+        item.closed_at = datetime.now(timezone.utc)
+
+    def list_taiga_open_tasks(
+        self,
+        project_slug: str | None = None,
+        limit: int = 20,
+    ) -> dict[str, Any]:
+        if not self.settings.taiga_configured:
+            raise ValueError("Taiga не настроена")
+
+        projects = self.list_projects()
+        if not projects:
+            projects = self.sync_taiga_projects()
+
+        if project_slug:
+            projects = [p for p in projects if p["slug"] == project_slug]
+            if not projects:
+                raise ValueError(
+                    f"Проект '{project_slug}' не найден. Вызови sync_taiga_projects."
+                )
+
+        client = TaigaClient()
+        blocks: list[dict[str, Any]] = []
+
+        for proj in projects:
+            stories = client.list_open_userstories(proj["taiga_id"], limit=limit)
+            tasks = client.list_open_tasks(proj["taiga_id"], limit=limit)
+            blocks.append(
+                {
+                    "slug": proj["slug"],
+                    "name": proj["name"],
+                    "taiga_id": proj["taiga_id"],
+                    "stories": [
+                        {
+                            "ref": s.get("ref"),
+                            "subject": s.get("subject", ""),
+                            "url": client.story_url(proj["taiga_id"], s.get("ref", 0)),
+                        }
+                        for s in stories
+                    ],
+                    "tasks": [
+                        {
+                            "ref": t.get("ref"),
+                            "subject": t.get("subject", ""),
+                            "user_story": t.get("user_story"),
+                        }
+                        for t in tasks
+                    ],
+                }
+            )
+
+        total_stories = sum(len(b["stories"]) for b in blocks)
+        total_tasks = sum(len(b["tasks"]) for b in blocks)
+        return {
+            "projects": blocks,
+            "total_stories": total_stories,
+            "total_tasks": total_tasks,
+        }
+
+    def list_work_items(self, limit: int = 30, status: str | None = None) -> list[dict[str, Any]]:
+        stmt = select(WorkItem).where(WorkItem.user_id == self.user_id).order_by(WorkItem.created_at.desc()).limit(limit)
+        if status:
+            stmt = stmt.where(WorkItem.status == status)
+        items = self.db.scalars(stmt).all()
+        settings = get_settings()
+        return [
+            {
+                "id": i.id,
+                "title": i.title,
+                "status": i.status,
+                "taiga_slug": i.taiga_slug,
+                "taiga_ref": i.taiga_story_ref,
+                "gitea_issue": i.gitea_issue_number,
+                "branch": i.suggested_branch,
+                "taiga_url": f"{settings.taiga_public_url}/project/0/{i.taiga_project_id}/us/{i.taiga_story_ref}",
+                "gitea_url": (
+                    f"{settings.gitea_public_url}/{i.gitea_owner}/{i.gitea_repo}/issues/{i.gitea_issue_number}"
+                    if i.gitea_issue_number
+                    else ""
+                ),
+                "created_at": i.created_at.isoformat() if i.created_at else None,
+            }
+            for i in items
+        ]

backend/app/rag/__init__.py

@@ -0,0 +1,5 @@
+"""RAG: embeddings, Qdrant store, retrieval, ingest."""
+
+from app.rag import chunker, embeddings, ingest, retriever, store
+
+__all__ = ["chunker", "embeddings", "ingest", "retriever", "store"]

backend/app/rag/chunker.py

@@ -0,0 +1,20 @@
+from __future__ import annotations
+
+
+def chunk_text(text: str, *, chunk_size: int = 800, overlap: int = 120) -> list[str]:
+    cleaned = (text or "").strip()
+    if not cleaned:
+        return []
+    if len(cleaned) <= chunk_size:
+        return [cleaned]
+    chunks: list[str] = []
+    start = 0
+    while start < len(cleaned):
+        end = min(len(cleaned), start + chunk_size)
+        piece = cleaned[start:end].strip()
+        if piece:
+            chunks.append(piece)
+        if end >= len(cleaned):
+            break
+        start = max(0, end - overlap)
+    return chunks

backend/app/rag/embeddings.py

@@ -0,0 +1,10 @@
+from __future__ import annotations
+
+from app.llm.client import LLMClient
+
+
+async def embed_texts(texts: list[str]) -> list[list[float]]:
+    if not texts:
+        return []
+    client = LLMClient()
+    return await client.embed(texts)

backend/app/rag/ingest.py

@@ -0,0 +1,152 @@
+from __future__ import annotations
+
+import hashlib
+from datetime import datetime, timezone
+from pathlib import Path
+from typing import Any
+
+from qdrant_client.http import models as qm
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.config import get_settings
+from app.db.models import ChatSession, Document, DocumentChunk, MemoryFact
+from app.rag import embeddings
+from app.rag.chunker import chunk_text
+from app.rag.store import (
+    COLLECTION_DOC_CHUNKS,
+    COLLECTION_FACTS,
+    COLLECTION_SUMMARIES,
+    delete_by_filter,
+    upsert_points,
+)
+
+
+async def index_memory_fact(fact: MemoryFact) -> None:
+    settings = get_settings()
+    if not settings.rag_enabled or not fact.active:
+        return
+    vectors = await embeddings.embed_texts([fact.content])
+    if not vectors:
+        return
+    upsert_points(
+        COLLECTION_FACTS,
+        [
+            qm.PointStruct(
+                id=int(fact.id),
+                vector=vectors[0],
+                payload={
+                    "user_id": fact.user_id,
+                    "fact_id": fact.id,
+                    "category": fact.category,
+                    "content": fact.content,
+                    "importance": fact.importance,
+                },
+            )
+        ],
+    )
+
+
+async def deactivate_memory_fact(fact_id: int) -> None:
+    settings = get_settings()
+    if not settings.rag_enabled:
+        return
+    delete_by_filter(
+        COLLECTION_FACTS,
+        [qm.FieldCondition(key="fact_id", match=qm.MatchValue(value=fact_id))],
+    )
+
+
+async def index_session_summary(session_id: int, summary: str) -> None:
+    settings = get_settings()
+    if not settings.rag_enabled or not summary.strip():
+        return
+    from app.db.base import SessionLocal
+
+    user_id = 1
+    db = SessionLocal()
+    try:
+        session = db.get(ChatSession, session_id)
+        if session:
+            user_id = session.user_id
+    finally:
+        db.close()
+
+    vectors = await embeddings.embed_texts([summary])
+    if not vectors:
+        return
+    upsert_points(
+        COLLECTION_SUMMARIES,
+        [
+            qm.PointStruct(
+                id=int(session_id),
+                vector=vectors[0],
+                payload={"user_id": user_id, "session_id": session_id, "summary": summary[:4000]},
+            )
+        ],
+    )
+
+
+async def ingest_document_file(
+    db: Session,
+    *,
+    user_id: int,
+    title: str,
+    filename: str,
+    raw_bytes: bytes,
+) -> dict[str, Any]:
+    settings = get_settings()
+    text = raw_bytes.decode("utf-8", errors="replace").strip()
+    if not text:
+        raise ValueError("Пустой документ")
+
+    digest = hashlib.sha256(raw_bytes).hexdigest()
+    doc = Document(
+        user_id=user_id,
+        title=title or filename,
+        filename=filename,
+        content_hash=digest,
+        size_bytes=len(raw_bytes),
+    )
+    db.add(doc)
+    db.flush()
+
+    chunks = chunk_text(text)
+    chunk_rows: list[DocumentChunk] = []
+    for idx, piece in enumerate(chunks):
+        row = DocumentChunk(document_id=doc.id, chunk_index=idx, content=piece)
+        db.add(row)
+        chunk_rows.append(row)
+    db.commit()
+    db.refresh(doc)
+
+    if settings.rag_enabled and chunks:
+        vectors = await embeddings.embed_texts(chunks)
+        points: list[qm.PointStruct] = []
+        for row, vector in zip(chunk_rows, vectors, strict=False):
+            db.refresh(row)
+            point_id = int(row.id)
+            points.append(
+                qm.PointStruct(
+                    id=point_id,
+                    vector=vector,
+                    payload={
+                        "user_id": user_id,
+                        "document_id": doc.id,
+                        "chunk_id": row.id,
+                        "chunk_index": row.chunk_index,
+                        "title": doc.title,
+                        "content": row.content,
+                    },
+                )
+            )
+        upsert_points(COLLECTION_DOC_CHUNKS, points)
+
+    return {
+        "id": doc.id,
+        "title": doc.title,
+        "filename": doc.filename,
+        "chunk_count": len(chunks),
+        "size_bytes": doc.size_bytes,
+        "created_at": doc.created_at.isoformat() if doc.created_at else None,
+    }

backend/app/rag/migrate_memory_to_qdrant.py

@@ -0,0 +1,37 @@
+# Migrate active memory facts into Qdrant
+from __future__ import annotations
+
+import asyncio
+
+from sqlalchemy import select
+
+from app.config import get_settings
+from app.db.base import SessionLocal, init_db
+from app.db.models import MemoryFact, SessionSummary
+from app.rag.ingest import index_memory_fact, index_session_summary
+from app.rag.store import ensure_collections
+
+
+async def main() -> None:
+    settings = get_settings()
+    if not settings.rag_enabled:
+        print("RAG disabled; set RAG_ENABLED=true")
+        return
+    init_db()
+    ensure_collections()
+    db = SessionLocal()
+    try:
+        facts = db.scalars(select(MemoryFact).where(MemoryFact.active.is_(True))).all()
+        for fact in facts:
+            await index_memory_fact(fact)
+        summaries = db.scalars(select(SessionSummary)).all()
+        for row in summaries:
+            if row.summary:
+                await index_session_summary(row.session_id, row.summary)
+        print(f"Indexed {len(facts)} facts and {len(summaries)} summaries")
+    finally:
+        db.close()
+
+
+if __name__ == "__main__":
+    asyncio.run(main())

backend/app/rag/retriever.py

@@ -0,0 +1,67 @@
+from __future__ import annotations
+
+from typing import Any
+
+from qdrant_client.http import models as qm
+
+from app.config import get_settings
+from app.rag import embeddings
+from app.rag.store import COLLECTION_DOC_CHUNKS, COLLECTION_FACTS, search
+
+
+def _user_filter(user_id: int) -> qm.Filter:
+    return qm.Filter(
+        must=[qm.FieldCondition(key="user_id", match=qm.MatchValue(value=user_id))]
+    )
+
+
+async def retrieve_memory_facts(
+    query: str, *, user_id: int, top_k: int | None = None
+) -> list[dict[str, Any]]:
+    settings = get_settings()
+    if not settings.rag_enabled or not query.strip():
+        return []
+    k = top_k or settings.rag_top_k
+    vectors = await embeddings.embed_texts([query])
+    if not vectors:
+        return []
+    hits = search(COLLECTION_FACTS, vectors[0], limit=k, query_filter=_user_filter(user_id))
+    results: list[dict[str, Any]] = []
+    for hit in hits:
+        payload = hit.payload or {}
+        results.append(
+            {
+                "id": payload.get("fact_id") or hit.id,
+                "category": payload.get("category", "fact"),
+                "content": payload.get("content", ""),
+                "score": hit.score,
+            }
+        )
+    return results
+
+
+async def retrieve_document_chunks(
+    query: str, *, user_id: int, top_k: int = 6
+) -> list[dict[str, Any]]:
+    settings = get_settings()
+    if not settings.rag_enabled or not query.strip():
+        return []
+    vectors = await embeddings.embed_texts([query])
+    if not vectors:
+        return []
+    hits = search(
+        COLLECTION_DOC_CHUNKS, vectors[0], limit=top_k, query_filter=_user_filter(user_id)
+    )
+    out: list[dict[str, Any]] = []
+    for hit in hits:
+        payload = hit.payload or {}
+        out.append(
+            {
+                "document_id": payload.get("document_id"),
+                "chunk_index": payload.get("chunk_index"),
+                "title": payload.get("title", ""),
+                "content": payload.get("content", ""),
+                "score": hit.score,
+            }
+        )
+    return out

backend/app/rag/store.py

@@ -0,0 +1,64 @@
+from __future__ import annotations
+
+import logging
+from typing import Any
+
+from qdrant_client import QdrantClient
+from qdrant_client.http import models as qm
+
+from app.config import get_settings
+
+logger = logging.getLogger(__name__)
+
+COLLECTION_FACTS = "memory_facts"
+COLLECTION_SUMMARIES = "session_summaries"
+COLLECTION_DOC_CHUNKS = "document_chunks"
+VECTOR_SIZE = 1536
+
+
+def _client() -> QdrantClient:
+    settings = get_settings()
+    return QdrantClient(url=settings.qdrant_url)
+
+
+def ensure_collections() -> None:
+    settings = get_settings()
+    if not settings.rag_enabled:
+        return
+    client = _client()
+    for name in (COLLECTION_FACTS, COLLECTION_SUMMARIES, COLLECTION_DOC_CHUNKS):
+        if client.collection_exists(name):
+            continue
+        client.create_collection(
+            collection_name=name,
+            vectors_config=qm.VectorParams(size=VECTOR_SIZE, distance=qm.Distance.COSINE),
+        )
+        logger.info("Created Qdrant collection %s", name)
+
+
+def upsert_points(collection: str, points: list[qm.PointStruct]) -> None:
+    if not points:
+        return
+    _client().upsert(collection_name=collection, points=points)
+
+
+def delete_by_filter(collection: str, must: list[qm.FieldCondition]) -> None:
+    _client().delete(
+        collection_name=collection,
+        points_selector=qm.FilterSelector(filter=qm.Filter(must=must)),
+    )
+
+
+def search(
+    collection: str,
+    vector: list[float],
+    *,
+    limit: int,
+    query_filter: qm.Filter | None = None,
+) -> list[qm.ScoredPoint]:
+    return _client().search(
+        collection_name=collection,
+        query_vector=vector,
+        limit=limit,
+        query_filter=query_filter,
+    )

backend/app/reminders/completion.py

@@ -1,73 +1,74 @@
-import logging
-from datetime import datetime, timezone
-
-from sqlalchemy.orm import Session
-
-from app.character.service import CharacterService
-from app.chat.notice_inbox import post_character_comment_to_latest_chat, post_notice_to_latest_chat
-from app.db.models import Reminder
-from app.llm.client import LLMClient
-from app.reminders.service import RECURRENCE_NONE, _advance_due, _format_local
-
-logger = logging.getLogger(__name__)
-
-
-def format_reminder_notice(row: Reminder) -> str:
-    local_when = _format_local(row.due_at, row.timezone, all_day=row.all_day)
-    notice = f"📅 **Напоминание** · {row.title}\n\n_{local_when}_"
-    if row.notes:
-        notice += f"\n{row.notes}"
-    return notice
-
-
-class ReminderCompletionHandler:
-    def __init__(self, db: Session):
-        self.db = db
-        self.llm = LLMClient()
-        self.character = CharacterService()
-
-    async def _generate_llm_comment(self, row: Reminder, local_when: str) -> str:
-        notes_part = f"\nЗаметки: {row.notes}" if row.notes else ""
-        rec_part = ""
-        if row.recurrence and row.recurrence != RECURRENCE_NONE:
-            rec_part = f"\nПовтор: {row.recurrence}"
-
-        system = self.character.get_system_prompt()
-        user_prompt = f"""Сработало напоминание.
-Заголовок: {row.title}
-Время: {local_when}{notes_part}{rec_part}
-
-Напиши пользователю короткое сообщение (2-4 предложения) на русском: напомни о деле, поддержи или предложи действие. Без markdown и без эмодзи."""
-
-        result = await self.llm.complete(
-            [
-                {"role": "system", "content": system},
-                {"role": "user", "content": user_prompt},
-            ],
-            temperature=0.8,
-            visible_reply=True,
-        )
-        return (result.get("content") or "").strip() or f"Напоминание: {row.title}"
-
-    def _mark_fired(self, row: Reminder, now: datetime) -> None:
-        row.last_fired_at = now
-        if row.recurrence == RECURRENCE_NONE:
-            row.completed_at = now
-            row.enabled = False
-        else:
-            row.due_at = _advance_due(row.due_at, row.recurrence)
-            row.last_fired_at = None
-        row.updated_at = now
-
-    async def process(self, row: Reminder) -> None:
-        local_when = _format_local(row.due_at, row.timezone, all_day=row.all_day)
-        post_notice_to_latest_chat(format_reminder_notice(row))
-
-        try:
-            comment = await self._generate_llm_comment(row, local_when)
-            if comment:
-                post_character_comment_to_latest_chat(comment)
-        except Exception:
-            logger.exception("Reminder LLM comment failed (id=%s)", row.id)
-
-        self._mark_fired(row, datetime.now(timezone.utc))
+import logging
+from datetime import datetime, timezone
+
+from sqlalchemy.orm import Session
+
+from app.character.service import CharacterService
+from app.chat.notice_inbox import post_character_comment_to_latest_chat, post_notice_to_latest_chat
+from app.db.models import Reminder
+from app.llm.client import LLMClient
+from app.reminders.service import RECURRENCE_NONE, _advance_due, _format_local
+
+logger = logging.getLogger(__name__)
+
+
+def format_reminder_notice(row: Reminder) -> str:
+    local_when = _format_local(row.due_at, row.timezone, all_day=row.all_day)
+    notice = f"📅 **Напоминание** · {row.title}\n\n_{local_when}_"
+    if row.notes:
+        notice += f"\n{row.notes}"
+    return notice
+
+
+class ReminderCompletionHandler:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+        self.llm = LLMClient()
+        self.character = CharacterService(db, user_id)
+
+    async def _generate_llm_comment(self, row: Reminder, local_when: str) -> str:
+        notes_part = f"\nЗаметки: {row.notes}" if row.notes else ""
+        rec_part = ""
+        if row.recurrence and row.recurrence != RECURRENCE_NONE:
+            rec_part = f"\nПовтор: {row.recurrence}"
+
+        system = self.character.get_system_prompt()
+        user_prompt = f"""Сработало напоминание.
+Заголовок: {row.title}
+Время: {local_when}{notes_part}{rec_part}
+
+Напиши пользователю короткое сообщение (2-4 предложения) на русском: напомни о деле, поддержи или предложи действие. Без markdown и без эмодзи."""
+
+        result = await self.llm.complete(
+            [
+                {"role": "system", "content": system},
+                {"role": "user", "content": user_prompt},
+            ],
+            temperature=0.8,
+            visible_reply=True,
+        )
+        return (result.get("content") or "").strip() or f"Напоминание: {row.title}"
+
+    def _mark_fired(self, row: Reminder, now: datetime) -> None:
+        row.last_fired_at = now
+        if row.recurrence == RECURRENCE_NONE:
+            row.completed_at = now
+            row.enabled = False
+        else:
+            row.due_at = _advance_due(row.due_at, row.recurrence)
+            row.last_fired_at = None
+        row.updated_at = now
+
+    async def process(self, row: Reminder) -> None:
+        local_when = _format_local(row.due_at, row.timezone, all_day=row.all_day)
+        post_notice_to_latest_chat(format_reminder_notice(row), self.user_id)
+
+        try:
+            comment = await self._generate_llm_comment(row, local_when)
+            if comment:
+                post_character_comment_to_latest_chat(comment, self.user_id)
+        except Exception:
+            logger.exception("Reminder LLM comment failed (id=%s)", row.id)
+
+        self._mark_fired(row, datetime.now(timezone.utc))

backend/app/reminders/context.py

@@ -1,33 +1,33 @@
-from typing import Any
-
-from sqlalchemy.orm import Session
-
-from app.reminders.service import RemindersService
-
-MAX_IN_CONTEXT = 10
-
-
-def get_reminders_snapshot(db: Session) -> dict[str, Any]:
-    return RemindersService(db).snapshot()
-
-
-def format_reminders_context(snapshot: dict[str, Any]) -> str:
-    lines = ["[Напоминания]"]
-    upcoming = snapshot.get("upcoming") or []
-    tz = snapshot.get("timezone", "Europe/Moscow")
-
-    if not upcoming:
-        lines.append(
-            "Ближайших напоминаний нет. "
-            "create_reminder для «напомни через 15 минут», «завтра утром», точной даты."
-        )
-        return "\n".join(lines)
-
-    lines.append(f"Часовой пояс: {tz}. Tools: list_reminders, create_reminder, update_reminder, delete_reminder, complete_reminder.")
-    for item in upcoming[:MAX_IN_CONTEXT]:
-        rec = item.get("recurrence", "none")
-        rec_label = f" · повтор: {rec}" if rec and rec != "none" else ""
-        lines.append(
-            f"- #{item['id']} **{item['title']}** · {item.get('due_at_local', item.get('due_at'))}{rec_label}"
-        )
-    return "\n".join(lines)
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.reminders.service import RemindersService
+
+MAX_IN_CONTEXT = 10
+
+
+def get_reminders_snapshot(db: Session, user_id: int) -> dict[str, Any]:
+    return RemindersService(db, user_id).snapshot()
+
+
+def format_reminders_context(snapshot: dict[str, Any]) -> str:
+    lines = ["[Напоминания]"]
+    upcoming = snapshot.get("upcoming") or []
+    tz = snapshot.get("timezone", "Europe/Moscow")
+
+    if not upcoming:
+        lines.append(
+            "Ближайших напоминаний нет. "
+            "create_reminder для «напомни через 15 минут», «завтра утром», точной даты."
+        )
+        return "\n".join(lines)
+
+    lines.append(f"Часовой пояс: {tz}. Tools: list_reminders, create_reminder, update_reminder, delete_reminder, complete_reminder.")
+    for item in upcoming[:MAX_IN_CONTEXT]:
+        rec = item.get("recurrence", "none")
+        rec_label = f" · повтор: {rec}" if rec and rec != "none" else ""
+        lines.append(
+            f"- #{item['id']} **{item['title']}** · {item.get('due_at_local', item.get('due_at'))}{rec_label}"
+        )
+    return "\n".join(lines)

backend/app/reminders/fire.py

@@ -1,45 +1,50 @@
-import logging
-from datetime import datetime, timezone
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.db.models import Reminder
-from app.reminders.completion import ReminderCompletionHandler
-from app.reminders.notify import bump_notify_seq
-
-logger = logging.getLogger(__name__)
-
-
-def _utcnow() -> datetime:
-    return datetime.now(timezone.utc)
-
-
-def get_due_reminders(db: Session) -> list[Reminder]:
-    now = _utcnow()
-    stmt = (
-        select(Reminder)
-        .where(
-            Reminder.enabled.is_(True),
-            Reminder.completed_at.is_(None),
-            Reminder.due_at <= now,
-        )
-        .order_by(Reminder.due_at.asc())
-    )
-    rows = list(db.scalars(stmt).all())
-    return [row for row in rows if not (row.last_fired_at and row.last_fired_at >= row.due_at)]
-
-
-async def process_due_reminders(db: Session) -> int:
-    due = get_due_reminders(db)
-    if not due:
-        return 0
-
-    handler = ReminderCompletionHandler(db)
-    for row in due:
-        await handler.process(row)
-
-    db.commit()
-    bump_notify_seq(db)
-    logger.info("Reminders fired: %d", len(due))
-    return len(due)
+import logging
+from datetime import datetime, timezone
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import Reminder, User
+from app.reminders.completion import ReminderCompletionHandler
+from app.reminders.notify import bump_notify_seq
+
+logger = logging.getLogger(__name__)
+
+
+def _utcnow() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+def get_due_reminders(db: Session, user_id: int) -> list[Reminder]:
+    now = _utcnow()
+    stmt = (
+        select(Reminder)
+        .where(
+            Reminder.user_id == user_id,
+            Reminder.enabled.is_(True),
+            Reminder.completed_at.is_(None),
+            Reminder.due_at <= now,
+        )
+        .order_by(Reminder.due_at.asc())
+    )
+    rows = list(db.scalars(stmt).all())
+    return [row for row in rows if not (row.last_fired_at and row.last_fired_at >= row.due_at)]
+
+
+async def process_due_reminders(db: Session) -> int:
+    users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+    total = 0
+    for user in users:
+        due = get_due_reminders(db, user.id)
+        if not due:
+            continue
+        handler = ReminderCompletionHandler(db, user.id)
+        for row in due:
+            await handler.process(row)
+        total += len(due)
+
+    if total:
+        db.commit()
+        bump_notify_seq(db)
+        logger.info("Reminders fired: %d", total)
+    return total

backend/app/reminders/service.py

@@ -1,237 +1,245 @@
-import calendar
-from datetime import datetime, timedelta, timezone
-from typing import Any
-from zoneinfo import ZoneInfo
-
-from sqlalchemy import select
-from sqlalchemy.orm import Session
-
-from app.db.models import Reminder
-from app.homelab.context import resolve_timezone
-from app.reminders.notify import bump_notify_seq, get_notify_seq
-
-RECURRENCE_NONE = "none"
-RECURRENCE_DAILY = "daily"
-RECURRENCE_WEEKLY = "weekly"
-RECURRENCE_MONTHLY = "monthly"
-RECURRENCE_YEARLY = "yearly"
-VALID_RECURRENCE = frozenset({
-    RECURRENCE_NONE,
-    RECURRENCE_DAILY,
-    RECURRENCE_WEEKLY,
-    RECURRENCE_MONTHLY,
-    RECURRENCE_YEARLY,
-})
-
-
-def _utcnow() -> datetime:
-    return datetime.now(timezone.utc)
-
-
-def _parse_due_at(raw: str, tz_name: str) -> datetime:
-    clean = raw.strip()
-    if not clean:
-        raise ValueError("due_at не может быть пустым")
-    try:
-        dt = datetime.fromisoformat(clean.replace("Z", "+00:00"))
-    except ValueError as exc:
-        raise ValueError(f"Неверный формат даты: {raw}") from exc
-    if dt.tzinfo is None:
-        try:
-            dt = dt.replace(tzinfo=ZoneInfo(tz_name))
-        except Exception:
-            dt = dt.replace(tzinfo=ZoneInfo("Europe/Moscow"))
-    return dt.astimezone(timezone.utc)
-
-
-def _advance_due(due_at: datetime, recurrence: str) -> datetime:
-    if recurrence == RECURRENCE_DAILY:
-        return due_at + timedelta(days=1)
-    if recurrence == RECURRENCE_WEEKLY:
-        return due_at + timedelta(weeks=1)
-    if recurrence == RECURRENCE_MONTHLY:
-        month = due_at.month + 1
-        year = due_at.year
-        if month > 12:
-            month = 1
-            year += 1
-        day = min(due_at.day, calendar.monthrange(year, month)[1])
-        return due_at.replace(year=year, month=month, day=day)
-    if recurrence == RECURRENCE_YEARLY:
-        year = due_at.year + 1
-        day = min(due_at.day, calendar.monthrange(year, due_at.month)[1])
-        return due_at.replace(year=year, day=day)
-    return due_at
-
-
-def _format_local(dt: datetime, tz_name: str, *, all_day: bool = False) -> str:
-    try:
-        local = dt.astimezone(ZoneInfo(tz_name))
-    except Exception:
-        local = dt.astimezone(ZoneInfo("Europe/Moscow"))
-    if all_day:
-        return local.strftime("%Y-%m-%d")
-    return local.strftime("%Y-%m-%d %H:%M")
-
-
-class RemindersService:
-    def __init__(self, db: Session):
-        self.db = db
-
-    def _tz(self) -> str:
-        return resolve_timezone(self.db)
-
-    def _to_dict(self, row: Reminder) -> dict[str, Any]:
-        tz = row.timezone or self._tz()
-        return {
-            "id": row.id,
-            "title": row.title,
-            "notes": row.notes,
-            "due_at": row.due_at.isoformat(),
-            "due_at_local": _format_local(row.due_at, tz, all_day=row.all_day),
-            "all_day": row.all_day,
-            "recurrence": row.recurrence,
-            "enabled": row.enabled,
-            "completed_at": row.completed_at.isoformat() if row.completed_at else None,
-            "timezone": tz,
-            "created_at": row.created_at.isoformat() if row.created_at else None,
-        }
-
-    def snapshot(self) -> dict[str, Any]:
-        upcoming = self.list_upcoming(limit=12)
-        return {
-            "notify_seq": get_notify_seq(self.db),
-            "upcoming": upcoming,
-            "upcoming_count": len(upcoming),
-            "timezone": self._tz(),
-        }
-
-    def list_upcoming(self, *, limit: int = 30) -> list[dict[str, Any]]:
-        stmt = (
-            select(Reminder)
-            .where(
-                Reminder.enabled.is_(True),
-                Reminder.completed_at.is_(None),
-            )
-            .order_by(Reminder.due_at.asc())
-            .limit(limit)
-        )
-        return [self._to_dict(row) for row in self.db.scalars(stmt).all()]
-
-    def list_in_range(
-        self,
-        *,
-        date_from: datetime,
-        date_to: datetime,
-    ) -> list[dict[str, Any]]:
-        stmt = (
-            select(Reminder)
-            .where(
-                Reminder.enabled.is_(True),
-                Reminder.completed_at.is_(None),
-                Reminder.due_at >= date_from,
-                Reminder.due_at < date_to,
-            )
-            .order_by(Reminder.due_at.asc())
-        )
-        return [self._to_dict(row) for row in self.db.scalars(stmt).all()]
-
-    def get(self, reminder_id: int) -> dict[str, Any] | None:
-        row = self.db.get(Reminder, reminder_id)
-        return self._to_dict(row) if row else None
-
-    def create(
-        self,
-        *,
-        title: str,
-        due_at: str,
-        notes: str = "",
-        all_day: bool = False,
-        recurrence: str = RECURRENCE_NONE,
-    ) -> dict[str, Any]:
-        clean_title = title.strip()
-        if not clean_title:
-            raise ValueError("Название напоминания не может быть пустым")
-        rec = (recurrence or RECURRENCE_NONE).strip().lower()
-        if rec not in VALID_RECURRENCE:
-            raise ValueError(f"recurrence должен быть один из: {', '.join(sorted(VALID_RECURRENCE))}")
-
-        tz = self._tz()
-        due = _parse_due_at(due_at, tz)
-        row = Reminder(
-            title=clean_title,
-            notes=notes.strip(),
-            due_at=due,
-            all_day=all_day,
-            recurrence=rec,
-            timezone=tz,
-        )
-        self.db.add(row)
-        self.db.commit()
-        self.db.refresh(row)
-        bump_notify_seq(self.db)
-        return {"ok": True, "reminder": self._to_dict(row), "created": True}
-
-    def update(
-        self,
-        reminder_id: int,
-        *,
-        title: str | None = None,
-        due_at: str | None = None,
-        notes: str | None = None,
-        all_day: bool | None = None,
-        recurrence: str | None = None,
-        enabled: bool | None = None,
-    ) -> dict[str, Any]:
-        row = self.db.get(Reminder, reminder_id)
-        if not row:
-            raise ValueError("Напоминание не найдено")
-
-        if title is not None:
-            clean = title.strip()
-            if not clean:
-                raise ValueError("Название не может быть пустым")
-            row.title = clean
-        if notes is not None:
-            row.notes = notes.strip()
-        if due_at is not None:
-            row.due_at = _parse_due_at(due_at, row.timezone or self._tz())
-            row.last_fired_at = None
-        if all_day is not None:
-            row.all_day = all_day
-        if recurrence is not None:
-            rec = recurrence.strip().lower()
-            if rec not in VALID_RECURRENCE:
-                raise ValueError(f"recurrence должен быть один из: {', '.join(sorted(VALID_RECURRENCE))}")
-            row.recurrence = rec
-        if enabled is not None:
-            row.enabled = enabled
-
-        row.updated_at = _utcnow()
-        self.db.commit()
-        self.db.refresh(row)
-        bump_notify_seq(self.db)
-        return {"ok": True, "reminder": self._to_dict(row)}
-
-    def delete(self, reminder_id: int) -> dict[str, Any]:
-        row = self.db.get(Reminder, reminder_id)
-        if not row:
-            raise ValueError("Напоминание не найдено")
-        title = row.title
-        self.db.delete(row)
-        self.db.commit()
-        bump_notify_seq(self.db)
-        return {"ok": True, "deleted_id": reminder_id, "title": title}
-
-    def complete(self, reminder_id: int) -> dict[str, Any]:
-        row = self.db.get(Reminder, reminder_id)
-        if not row:
-            raise ValueError("Напоминание не найдено")
-        now = _utcnow()
-        row.completed_at = now
-        row.enabled = False
-        row.updated_at = now
-        self.db.commit()
-        self.db.refresh(row)
-        bump_notify_seq(self.db)
-        return {"ok": True, "reminder": self._to_dict(row)}
+import calendar
+from datetime import datetime, timedelta, timezone
+from typing import Any
+from zoneinfo import ZoneInfo
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import Reminder
+from app.memory.service import MemoryService
+from app.reminders.notify import bump_notify_seq, get_notify_seq
+
+RECURRENCE_NONE = "none"
+RECURRENCE_DAILY = "daily"
+RECURRENCE_WEEKLY = "weekly"
+RECURRENCE_MONTHLY = "monthly"
+RECURRENCE_YEARLY = "yearly"
+VALID_RECURRENCE = frozenset({
+    RECURRENCE_NONE,
+    RECURRENCE_DAILY,
+    RECURRENCE_WEEKLY,
+    RECURRENCE_MONTHLY,
+    RECURRENCE_YEARLY,
+})
+
+
+def _utcnow() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+def _parse_due_at(raw: str, tz_name: str) -> datetime:
+    clean = raw.strip()
+    if not clean:
+        raise ValueError("due_at не может быть пустым")
+    try:
+        dt = datetime.fromisoformat(clean.replace("Z", "+00:00"))
+    except ValueError as exc:
+        raise ValueError(f"Неверный формат даты: {raw}") from exc
+    if dt.tzinfo is None:
+        try:
+            dt = dt.replace(tzinfo=ZoneInfo(tz_name))
+        except Exception:
+            dt = dt.replace(tzinfo=ZoneInfo("Europe/Moscow"))
+    return dt.astimezone(timezone.utc)
+
+
+def _advance_due(due_at: datetime, recurrence: str) -> datetime:
+    if recurrence == RECURRENCE_DAILY:
+        return due_at + timedelta(days=1)
+    if recurrence == RECURRENCE_WEEKLY:
+        return due_at + timedelta(weeks=1)
+    if recurrence == RECURRENCE_MONTHLY:
+        month = due_at.month + 1
+        year = due_at.year
+        if month > 12:
+            month = 1
+            year += 1
+        day = min(due_at.day, calendar.monthrange(year, month)[1])
+        return due_at.replace(year=year, month=month, day=day)
+    if recurrence == RECURRENCE_YEARLY:
+        year = due_at.year + 1
+        day = min(due_at.day, calendar.monthrange(year, due_at.month)[1])
+        return due_at.replace(year=year, day=day)
+    return due_at
+
+
+def _format_local(dt: datetime, tz_name: str, *, all_day: bool = False) -> str:
+    try:
+        local = dt.astimezone(ZoneInfo(tz_name))
+    except Exception:
+        local = dt.astimezone(ZoneInfo("Europe/Moscow"))
+    if all_day:
+        return local.strftime("%Y-%m-%d")
+    return local.strftime("%Y-%m-%d %H:%M")
+
+
+class RemindersService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+
+    def _tz(self) -> str:
+        profile = MemoryService(self.db, self.user_id).get_profile()
+        tz = (profile.get("timezone") or "").strip()
+        return tz or "Europe/Moscow"
+
+    def _to_dict(self, row: Reminder) -> dict[str, Any]:
+        tz = row.timezone or self._tz()
+        return {
+            "id": row.id,
+            "title": row.title,
+            "notes": row.notes,
+            "due_at": row.due_at.isoformat(),
+            "due_at_local": _format_local(row.due_at, tz, all_day=row.all_day),
+            "all_day": row.all_day,
+            "recurrence": row.recurrence,
+            "enabled": row.enabled,
+            "completed_at": row.completed_at.isoformat() if row.completed_at else None,
+            "timezone": tz,
+            "created_at": row.created_at.isoformat() if row.created_at else None,
+        }
+
+    def snapshot(self) -> dict[str, Any]:
+        upcoming = self.list_upcoming(limit=12)
+        return {
+            "notify_seq": get_notify_seq(self.db),
+            "upcoming": upcoming,
+            "upcoming_count": len(upcoming),
+            "timezone": self._tz(),
+        }
+
+    def list_upcoming(self, *, limit: int = 30) -> list[dict[str, Any]]:
+        stmt = (
+            select(Reminder)
+            .where(
+                Reminder.user_id == self.user_id,
+                Reminder.enabled.is_(True),
+                Reminder.completed_at.is_(None),
+            )
+            .order_by(Reminder.due_at.asc())
+            .limit(limit)
+        )
+        return [self._to_dict(row) for row in self.db.scalars(stmt).all()]
+
+    def list_in_range(
+        self,
+        *,
+        date_from: datetime,
+        date_to: datetime,
+    ) -> list[dict[str, Any]]:
+        stmt = (
+            select(Reminder)
+            .where(
+                Reminder.user_id == self.user_id,
+                Reminder.enabled.is_(True),
+                Reminder.completed_at.is_(None),
+                Reminder.due_at >= date_from,
+                Reminder.due_at < date_to,
+            )
+            .order_by(Reminder.due_at.asc())
+        )
+        return [self._to_dict(row) for row in self.db.scalars(stmt).all()]
+
+    def get(self, reminder_id: int) -> dict[str, Any] | None:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            return None
+        return self._to_dict(row)
+
+    def create(
+        self,
+        *,
+        title: str,
+        due_at: str,
+        notes: str = "",
+        all_day: bool = False,
+        recurrence: str = RECURRENCE_NONE,
+    ) -> dict[str, Any]:
+        clean_title = title.strip()
+        if not clean_title:
+            raise ValueError("Название напоминания не может быть пустым")
+        rec = (recurrence or RECURRENCE_NONE).strip().lower()
+        if rec not in VALID_RECURRENCE:
+            raise ValueError(f"recurrence должен быть один из: {', '.join(sorted(VALID_RECURRENCE))}")
+
+        tz = self._tz()
+        due = _parse_due_at(due_at, tz)
+        row = Reminder(
+            user_id=self.user_id,
+            title=clean_title,
+            notes=notes.strip(),
+            due_at=due,
+            all_day=all_day,
+            recurrence=rec,
+            timezone=tz,
+        )
+        self.db.add(row)
+        self.db.commit()
+        self.db.refresh(row)
+        bump_notify_seq(self.db)
+        return {"ok": True, "reminder": self._to_dict(row), "created": True}
+
+    def update(
+        self,
+        reminder_id: int,
+        *,
+        title: str | None = None,
+        due_at: str | None = None,
+        notes: str | None = None,
+        all_day: bool | None = None,
+        recurrence: str | None = None,
+        enabled: bool | None = None,
+    ) -> dict[str, Any]:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            raise ValueError("Напоминание не найдено")
+
+        if title is not None:
+            clean = title.strip()
+            if not clean:
+                raise ValueError("Название не может быть пустым")
+            row.title = clean
+        if notes is not None:
+            row.notes = notes.strip()
+        if due_at is not None:
+            row.due_at = _parse_due_at(due_at, row.timezone or self._tz())
+            row.last_fired_at = None
+        if all_day is not None:
+            row.all_day = all_day
+        if recurrence is not None:
+            rec = recurrence.strip().lower()
+            if rec not in VALID_RECURRENCE:
+                raise ValueError(f"recurrence должен быть один из: {', '.join(sorted(VALID_RECURRENCE))}")
+            row.recurrence = rec
+        if enabled is not None:
+            row.enabled = enabled
+
+        row.updated_at = _utcnow()
+        self.db.commit()
+        self.db.refresh(row)
+        bump_notify_seq(self.db)
+        return {"ok": True, "reminder": self._to_dict(row)}
+
+    def delete(self, reminder_id: int) -> dict[str, Any]:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            raise ValueError("Напоминание не найдено")
+        title = row.title
+        self.db.delete(row)
+        self.db.commit()
+        bump_notify_seq(self.db)
+        return {"ok": True, "deleted_id": reminder_id, "title": title}
+
+    def complete(self, reminder_id: int) -> dict[str, Any]:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            raise ValueError("Напоминание не найдено")
+        now = _utcnow()
+        row.completed_at = now
+        row.enabled = False
+        row.updated_at = now
+        self.db.commit()
+        self.db.refresh(row)
+        bump_notify_seq(self.db)
+        return {"ok": True, "reminder": self._to_dict(row)}

backend/app/reminders_scoped/__init__.py

@@ -0,0 +1,3 @@
+from app.reminders_scoped.service import RemindersService
+
+__all__ = ["RemindersService"]

backend/app/reminders_scoped/completion.py

@@ -0,0 +1,74 @@
+import logging
+from datetime import datetime, timezone
+
+from sqlalchemy.orm import Session
+
+from app.character.service import CharacterService
+from app.chat.notice_inbox import post_character_comment_to_latest_chat, post_notice_to_latest_chat
+from app.db.models import Reminder
+from app.llm.client import LLMClient
+from app.reminders_scoped.service import RECURRENCE_NONE, _advance_due, _format_local
+
+logger = logging.getLogger(__name__)
+
+
+def format_reminder_notice(row: Reminder) -> str:
+    local_when = _format_local(row.due_at, row.timezone, all_day=row.all_day)
+    notice = f"📅 **Напоминание** · {row.title}\n\n_{local_when}_"
+    if row.notes:
+        notice += f"\n{row.notes}"
+    return notice
+
+
+class ReminderCompletionHandler:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+        self.llm = LLMClient()
+        self.character = CharacterService(db, user_id)
+
+    async def _generate_llm_comment(self, row: Reminder, local_when: str) -> str:
+        notes_part = f"\nЗаметки: {row.notes}" if row.notes else ""
+        rec_part = ""
+        if row.recurrence and row.recurrence != RECURRENCE_NONE:
+            rec_part = f"\nПовтор: {row.recurrence}"
+
+        system = self.character.get_system_prompt()
+        user_prompt = f"""Сработало напоминание.
+Заголовок: {row.title}
+Время: {local_when}{notes_part}{rec_part}
+
+Напиши пользователю короткое сообщение (2-4 предложения) на русском: напомни о деле, поддержи или предложи действие. Без markdown и без эмодзи."""
+
+        result = await self.llm.complete(
+            [
+                {"role": "system", "content": system},
+                {"role": "user", "content": user_prompt},
+            ],
+            temperature=0.8,
+            visible_reply=True,
+        )
+        return (result.get("content") or "").strip() or f"Напоминание: {row.title}"
+
+    def _mark_fired(self, row: Reminder, now: datetime) -> None:
+        row.last_fired_at = now
+        if row.recurrence == RECURRENCE_NONE:
+            row.completed_at = now
+            row.enabled = False
+        else:
+            row.due_at = _advance_due(row.due_at, row.recurrence)
+            row.last_fired_at = None
+        row.updated_at = now
+
+    async def process(self, row: Reminder) -> None:
+        local_when = _format_local(row.due_at, row.timezone, all_day=row.all_day)
+        post_notice_to_latest_chat(format_reminder_notice(row), self.user_id)
+
+        try:
+            comment = await self._generate_llm_comment(row, local_when)
+            if comment:
+                post_character_comment_to_latest_chat(comment, self.user_id)
+        except Exception:
+            logger.exception("Reminder LLM comment failed (id=%s)", row.id)
+
+        self._mark_fired(row, datetime.now(timezone.utc))

backend/app/reminders_scoped/context.py

@@ -0,0 +1,33 @@
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.reminders_scoped.service import RemindersService
+
+MAX_IN_CONTEXT = 10
+
+
+def get_reminders_snapshot(db: Session, user_id: int) -> dict[str, Any]:
+    return RemindersService(db, user_id).snapshot()
+
+
+def format_reminders_context(snapshot: dict[str, Any]) -> str:
+    lines = ["[Напоминания]"]
+    upcoming = snapshot.get("upcoming") or []
+    tz = snapshot.get("timezone", "Europe/Moscow")
+
+    if not upcoming:
+        lines.append(
+            "Ближайших напоминаний нет. "
+            "create_reminder для «напомни через 15 минут», «завтра утром», точной даты."
+        )
+        return "\n".join(lines)
+
+    lines.append(f"Часовой пояс: {tz}. Tools: list_reminders, create_reminder, update_reminder, delete_reminder, complete_reminder.")
+    for item in upcoming[:MAX_IN_CONTEXT]:
+        rec = item.get("recurrence", "none")
+        rec_label = f" · повтор: {rec}" if rec and rec != "none" else ""
+        lines.append(
+            f"- #{item['id']} **{item['title']}** · {item.get('due_at_local', item.get('due_at'))}{rec_label}"
+        )
+    return "\n".join(lines)

backend/app/reminders_scoped/fire.py

@@ -0,0 +1,50 @@
+import logging
+from datetime import datetime, timezone
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import Reminder, User
+from app.reminders_scoped.completion import ReminderCompletionHandler
+from app.reminders.notify import bump_notify_seq
+
+logger = logging.getLogger(__name__)
+
+
+def _utcnow() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+def get_due_reminders(db: Session, user_id: int) -> list[Reminder]:
+    now = _utcnow()
+    stmt = (
+        select(Reminder)
+        .where(
+            Reminder.user_id == user_id,
+            Reminder.enabled.is_(True),
+            Reminder.completed_at.is_(None),
+            Reminder.due_at <= now,
+        )
+        .order_by(Reminder.due_at.asc())
+    )
+    rows = list(db.scalars(stmt).all())
+    return [row for row in rows if not (row.last_fired_at and row.last_fired_at >= row.due_at)]
+
+
+async def process_due_reminders(db: Session) -> int:
+    users = db.scalars(select(User).where(User.is_active.is_(True))).all()
+    total = 0
+    for user in users:
+        due = get_due_reminders(db, user.id)
+        if not due:
+            continue
+        handler = ReminderCompletionHandler(db, user.id)
+        for row in due:
+            await handler.process(row)
+        total += len(due)
+
+    if total:
+        db.commit()
+        bump_notify_seq(db)
+        logger.info("Reminders fired: %d", total)
+    return total

backend/app/reminders_scoped/service.py

@@ -0,0 +1,245 @@
+import calendar
+from datetime import datetime, timedelta, timezone
+from typing import Any
+from zoneinfo import ZoneInfo
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.db.models import Reminder
+from app.memory.service import MemoryService
+from app.reminders.notify import bump_notify_seq, get_notify_seq
+
+RECURRENCE_NONE = "none"
+RECURRENCE_DAILY = "daily"
+RECURRENCE_WEEKLY = "weekly"
+RECURRENCE_MONTHLY = "monthly"
+RECURRENCE_YEARLY = "yearly"
+VALID_RECURRENCE = frozenset({
+    RECURRENCE_NONE,
+    RECURRENCE_DAILY,
+    RECURRENCE_WEEKLY,
+    RECURRENCE_MONTHLY,
+    RECURRENCE_YEARLY,
+})
+
+
+def _utcnow() -> datetime:
+    return datetime.now(timezone.utc)
+
+
+def _parse_due_at(raw: str, tz_name: str) -> datetime:
+    clean = raw.strip()
+    if not clean:
+        raise ValueError("due_at не может быть пустым")
+    try:
+        dt = datetime.fromisoformat(clean.replace("Z", "+00:00"))
+    except ValueError as exc:
+        raise ValueError(f"Неверный формат даты: {raw}") from exc
+    if dt.tzinfo is None:
+        try:
+            dt = dt.replace(tzinfo=ZoneInfo(tz_name))
+        except Exception:
+            dt = dt.replace(tzinfo=ZoneInfo("Europe/Moscow"))
+    return dt.astimezone(timezone.utc)
+
+
+def _advance_due(due_at: datetime, recurrence: str) -> datetime:
+    if recurrence == RECURRENCE_DAILY:
+        return due_at + timedelta(days=1)
+    if recurrence == RECURRENCE_WEEKLY:
+        return due_at + timedelta(weeks=1)
+    if recurrence == RECURRENCE_MONTHLY:
+        month = due_at.month + 1
+        year = due_at.year
+        if month > 12:
+            month = 1
+            year += 1
+        day = min(due_at.day, calendar.monthrange(year, month)[1])
+        return due_at.replace(year=year, month=month, day=day)
+    if recurrence == RECURRENCE_YEARLY:
+        year = due_at.year + 1
+        day = min(due_at.day, calendar.monthrange(year, due_at.month)[1])
+        return due_at.replace(year=year, day=day)
+    return due_at
+
+
+def _format_local(dt: datetime, tz_name: str, *, all_day: bool = False) -> str:
+    try:
+        local = dt.astimezone(ZoneInfo(tz_name))
+    except Exception:
+        local = dt.astimezone(ZoneInfo("Europe/Moscow"))
+    if all_day:
+        return local.strftime("%Y-%m-%d")
+    return local.strftime("%Y-%m-%d %H:%M")
+
+
+class RemindersService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+
+    def _tz(self) -> str:
+        profile = MemoryService(self.db, self.user_id).get_profile()
+        tz = (profile.get("timezone") or "").strip()
+        return tz or "Europe/Moscow"
+
+    def _to_dict(self, row: Reminder) -> dict[str, Any]:
+        tz = row.timezone or self._tz()
+        return {
+            "id": row.id,
+            "title": row.title,
+            "notes": row.notes,
+            "due_at": row.due_at.isoformat(),
+            "due_at_local": _format_local(row.due_at, tz, all_day=row.all_day),
+            "all_day": row.all_day,
+            "recurrence": row.recurrence,
+            "enabled": row.enabled,
+            "completed_at": row.completed_at.isoformat() if row.completed_at else None,
+            "timezone": tz,
+            "created_at": row.created_at.isoformat() if row.created_at else None,
+        }
+
+    def snapshot(self) -> dict[str, Any]:
+        upcoming = self.list_upcoming(limit=12)
+        return {
+            "notify_seq": get_notify_seq(self.db),
+            "upcoming": upcoming,
+            "upcoming_count": len(upcoming),
+            "timezone": self._tz(),
+        }
+
+    def list_upcoming(self, *, limit: int = 30) -> list[dict[str, Any]]:
+        stmt = (
+            select(Reminder)
+            .where(
+                Reminder.user_id == self.user_id,
+                Reminder.enabled.is_(True),
+                Reminder.completed_at.is_(None),
+            )
+            .order_by(Reminder.due_at.asc())
+            .limit(limit)
+        )
+        return [self._to_dict(row) for row in self.db.scalars(stmt).all()]
+
+    def list_in_range(
+        self,
+        *,
+        date_from: datetime,
+        date_to: datetime,
+    ) -> list[dict[str, Any]]:
+        stmt = (
+            select(Reminder)
+            .where(
+                Reminder.user_id == self.user_id,
+                Reminder.enabled.is_(True),
+                Reminder.completed_at.is_(None),
+                Reminder.due_at >= date_from,
+                Reminder.due_at < date_to,
+            )
+            .order_by(Reminder.due_at.asc())
+        )
+        return [self._to_dict(row) for row in self.db.scalars(stmt).all()]
+
+    def get(self, reminder_id: int) -> dict[str, Any] | None:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            return None
+        return self._to_dict(row)
+
+    def create(
+        self,
+        *,
+        title: str,
+        due_at: str,
+        notes: str = "",
+        all_day: bool = False,
+        recurrence: str = RECURRENCE_NONE,
+    ) -> dict[str, Any]:
+        clean_title = title.strip()
+        if not clean_title:
+            raise ValueError("Название напоминания не может быть пустым")
+        rec = (recurrence or RECURRENCE_NONE).strip().lower()
+        if rec not in VALID_RECURRENCE:
+            raise ValueError(f"recurrence должен быть один из: {', '.join(sorted(VALID_RECURRENCE))}")
+
+        tz = self._tz()
+        due = _parse_due_at(due_at, tz)
+        row = Reminder(
+            user_id=self.user_id,
+            title=clean_title,
+            notes=notes.strip(),
+            due_at=due,
+            all_day=all_day,
+            recurrence=rec,
+            timezone=tz,
+        )
+        self.db.add(row)
+        self.db.commit()
+        self.db.refresh(row)
+        bump_notify_seq(self.db)
+        return {"ok": True, "reminder": self._to_dict(row), "created": True}
+
+    def update(
+        self,
+        reminder_id: int,
+        *,
+        title: str | None = None,
+        due_at: str | None = None,
+        notes: str | None = None,
+        all_day: bool | None = None,
+        recurrence: str | None = None,
+        enabled: bool | None = None,
+    ) -> dict[str, Any]:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            raise ValueError("Напоминание не найдено")
+
+        if title is not None:
+            clean = title.strip()
+            if not clean:
+                raise ValueError("Название не может быть пустым")
+            row.title = clean
+        if notes is not None:
+            row.notes = notes.strip()
+        if due_at is not None:
+            row.due_at = _parse_due_at(due_at, row.timezone or self._tz())
+            row.last_fired_at = None
+        if all_day is not None:
+            row.all_day = all_day
+        if recurrence is not None:
+            rec = recurrence.strip().lower()
+            if rec not in VALID_RECURRENCE:
+                raise ValueError(f"recurrence должен быть один из: {', '.join(sorted(VALID_RECURRENCE))}")
+            row.recurrence = rec
+        if enabled is not None:
+            row.enabled = enabled
+
+        row.updated_at = _utcnow()
+        self.db.commit()
+        self.db.refresh(row)
+        bump_notify_seq(self.db)
+        return {"ok": True, "reminder": self._to_dict(row)}
+
+    def delete(self, reminder_id: int) -> dict[str, Any]:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            raise ValueError("Напоминание не найдено")
+        title = row.title
+        self.db.delete(row)
+        self.db.commit()
+        bump_notify_seq(self.db)
+        return {"ok": True, "deleted_id": reminder_id, "title": title}
+
+    def complete(self, reminder_id: int) -> dict[str, Any]:
+        row = self.db.get(Reminder, reminder_id)
+        if not row or row.user_id != self.user_id:
+            raise ValueError("Напоминание не найдено")
+        now = _utcnow()
+        row.completed_at = now
+        row.enabled = False
+        row.updated_at = now
+        self.db.commit()
+        self.db.refresh(row)
+        bump_notify_seq(self.db)
+        return {"ok": True, "reminder": self._to_dict(row)}

backend/app/reminders_scoped/watcher.py

@@ -0,0 +1,31 @@
+import asyncio
+import logging
+
+from app.config import get_settings
+from app.db.base import SessionLocal
+from app.reminders_scoped.fire import process_due_reminders
+
+logger = logging.getLogger(__name__)
+
+WATCH_INTERVAL_SEC = 30
+
+
+async def reminders_watcher_loop() -> None:
+    while True:
+        try:
+            await asyncio.sleep(WATCH_INTERVAL_SEC)
+            if not get_settings().reminders_enabled:
+                continue
+            await _tick()
+        except asyncio.CancelledError:
+            raise
+        except Exception:
+            logger.exception("Reminders watcher error")
+
+
+async def _tick() -> None:
+    db = SessionLocal()
+    try:
+        await process_due_reminders(db)
+    finally:
+        db.close()

backend/app/settings/__init__.py

@@ -0,0 +1 @@
+"""Runtime settings stored in assistant_state."""

backend/app/settings/service.py

@@ -0,0 +1,98 @@
+from __future__ import annotations
+
+import json
+from datetime import datetime, timezone
+from typing import Any
+
+from sqlalchemy import select
+from sqlalchemy.orm import Session
+
+from app.config import Settings, get_settings
+from app.db.models import AssistantState
+
+SETTING_KEYS = (
+    "openrouter_model",
+    "memory_extract_model",
+    "openrouter_reasoning_effort",
+    "rag_enabled",
+    "rag_top_k",
+)
+
+
+class SettingsService:
+    def __init__(self, db: Session):
+        self.db = db
+        self._defaults = get_settings()
+
+    def _get_row(self, key: str) -> AssistantState | None:
+        return self.db.get(AssistantState, key)
+
+    def get_raw(self, key: str) -> str | None:
+        row = self._get_row(key)
+        if not row or not (row.value or "").strip():
+            return None
+        return row.value.strip()
+
+    def set_raw(self, key: str, value: str) -> None:
+        row = self._get_row(key)
+        if not row:
+            row = AssistantState(key=key, value=value)
+            self.db.add(row)
+        else:
+            row.value = value
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+
+    def _default_for(self, key: str) -> Any:
+        defaults: Settings = self._defaults
+        mapping = {
+            "openrouter_model": defaults.openrouter_model,
+            "memory_extract_model": defaults.memory_extract_model or defaults.openrouter_model,
+            "openrouter_reasoning_effort": defaults.openrouter_reasoning_effort,
+            "rag_enabled": defaults.rag_enabled,
+            "rag_top_k": defaults.rag_top_k,
+        }
+        return mapping[key]
+
+    def get_effective(self, key: str) -> Any:
+        raw = self.get_raw(key)
+        if raw is None:
+            return self._default_for(key)
+        if key == "rag_enabled":
+            return raw.lower() in ("1", "true", "yes", "on")
+        if key == "rag_top_k":
+            try:
+                return max(1, min(50, int(raw)))
+            except ValueError:
+                return self._default_for(key)
+        return raw
+
+    def snapshot(self) -> dict[str, Any]:
+        data: dict[str, Any] = {}
+        for key in SETTING_KEYS:
+            data[key] = self.get_effective(key)
+        data["embedding_model"] = self._defaults.embedding_model
+        data["memory_facts_in_context"] = self._defaults.memory_facts_in_context
+        data["qdrant_url"] = self._defaults.qdrant_url
+        return data
+
+    def patch(self, updates: dict[str, Any]) -> dict[str, Any]:
+        for key, value in updates.items():
+            if key not in SETTING_KEYS:
+                continue
+            if value is None:
+                row = self._get_row(key)
+                if row:
+                    self.db.delete(row)
+                    self.db.commit()
+                continue
+            if key == "rag_enabled":
+                stored = "true" if bool(value) else "false"
+            elif key == "rag_top_k":
+                stored = str(int(value))
+            else:
+                stored = str(value).strip()
+            if not stored and key != "rag_enabled":
+                continue
+            self.set_raw(key, stored)
+        return self.snapshot()

backend/app/shopping/context.py

@@ -1,47 +1,47 @@
-from typing import Any
-
-from sqlalchemy.orm import Session
-
-from app.shopping.service import ShoppingService
-
-MAX_LISTS_IN_CONTEXT = 8
-MAX_ITEMS_PER_LIST = 12
-
-
-def get_shopping_snapshot(db: Session) -> dict[str, Any]:
-    return ShoppingService(db).snapshot()
-
-
-def format_shopping_context(snapshot: dict[str, Any]) -> str:
-    lines = ["[Списки покупок]"]
-    lists = snapshot.get("lists") or []
-
-    if not lists:
-        lines.append("Списков пока нет. create_shopping_list или add_shopping_items.")
-        return "\n".join(lines)
-
-    lines.append(
-        f"Всего списков: {snapshot.get('list_count', len(lists))}, "
-        f"неотмеченных позиций: {snapshot.get('unchecked_items', 0)}."
-    )
-    lines.append("Для изменений вызывай tools: list_shopping_lists, add_shopping_items, check_shopping_item.")
-
-    for lst in lists[:MAX_LISTS_IN_CONTEXT]:
-        items = lst.get("items") or []
-        unchecked = [i for i in items if not i.get("checked")]
-        preview = unchecked[:MAX_ITEMS_PER_LIST]
-        parts = []
-        for item in preview:
-            qty = item.get("quantity")
-            unit = (item.get("unit") or "").strip()
-            label = item["text"]
-            if qty is not None:
-                label = f"{label} ({qty}{' ' + unit if unit else ''})"
-            parts.append(f"#{item['id']} {label}")
-        tail = f" +{len(unchecked) - len(preview)} ещё" if len(unchecked) > len(preview) else ""
-        if parts:
-            lines.append(f"- «{lst['name']}» (#{lst['id']}): {', '.join(parts)}{tail}")
-        else:
-            lines.append(f"- «{lst['name']}» (#{lst['id']}): всё отмечено или пусто")
-
-    return "\n".join(lines)
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.shopping.service import ShoppingService
+
+MAX_LISTS_IN_CONTEXT = 8
+MAX_ITEMS_PER_LIST = 12
+
+
+def get_shopping_snapshot(db: Session, user_id: int) -> dict[str, Any]:
+    return ShoppingService(db, user_id).snapshot()
+
+
+def format_shopping_context(snapshot: dict[str, Any]) -> str:
+    lines = ["[Списки покупок]"]
+    lists = snapshot.get("lists") or []
+
+    if not lists:
+        lines.append("Списков пока нет. create_shopping_list или add_shopping_items.")
+        return "\n".join(lines)
+
+    lines.append(
+        f"Всего списков: {snapshot.get('list_count', len(lists))}, "
+        f"неотмеченных позиций: {snapshot.get('unchecked_items', 0)}."
+    )
+    lines.append("Для изменений вызывай tools: list_shopping_lists, add_shopping_items, check_shopping_item.")
+
+    for lst in lists[:MAX_LISTS_IN_CONTEXT]:
+        items = lst.get("items") or []
+        unchecked = [i for i in items if not i.get("checked")]
+        preview = unchecked[:MAX_ITEMS_PER_LIST]
+        parts = []
+        for item in preview:
+            qty = item.get("quantity")
+            unit = (item.get("unit") or "").strip()
+            label = item["text"]
+            if qty is not None:
+                label = f"{label} ({qty}{' ' + unit if unit else ''})"
+            parts.append(f"#{item['id']} {label}")
+        tail = f" +{len(unchecked) - len(preview)} ещё" if len(unchecked) > len(preview) else ""
+        if parts:
+            lines.append(f"- «{lst['name']}» (#{lst['id']}): {', '.join(parts)}{tail}")
+        else:
+            lines.append(f"- «{lst['name']}» (#{lst['id']}): всё отмечено или пусто")
+
+    return "\n".join(lines)

backend/app/shopping/service.py

@@ -1,223 +1,224 @@
-from datetime import datetime, timezone
-from typing import Any
-
-from sqlalchemy import func, select
-from sqlalchemy.orm import Session, selectinload
-
-from app.db.models import ShoppingList, ShoppingListItem
-
-
-class ShoppingService:
-    def __init__(self, db: Session):
-        self.db = db
-
-    def snapshot(self) -> dict[str, Any]:
-        lists = self.list_lists(include_items=True)
-        total_items = sum(len(lst.get("items") or []) for lst in lists)
-        unchecked = sum(
-            1
-            for lst in lists
-            for item in (lst.get("items") or [])
-            if not item.get("checked")
-        )
-        return {
-            "lists": lists,
-            "list_count": len(lists),
-            "total_items": total_items,
-            "unchecked_items": unchecked,
-        }
-
-    def list_lists(self, *, include_items: bool = False) -> list[dict[str, Any]]:
-        stmt = select(ShoppingList).order_by(ShoppingList.sort_order, ShoppingList.name)
-        if include_items:
-            stmt = stmt.options(selectinload(ShoppingList.items))
-        rows = list(self.db.scalars(stmt).all())
-        return [self._list_to_dict(row, include_items=include_items) for row in rows]
-
-    def get_list(
-        self,
-        list_id: int | None = None,
-        *,
-        name: str | None = None,
-    ) -> dict[str, Any] | None:
-        row = self._resolve_list(list_id=list_id, name=name)
-        if not row:
-            return None
-        return self._list_to_dict(row, include_items=True)
-
-    def create_list(self, name: str) -> dict[str, Any]:
-        clean = name.strip()
-        if not clean:
-            raise ValueError("Название списка не может быть пустым")
-        existing = self.db.scalar(select(ShoppingList).where(ShoppingList.name == clean))
-        if existing:
-            return {"ok": True, "list": self._list_to_dict(existing, include_items=True), "created": False}
-
-        max_order = self.db.scalar(select(func.max(ShoppingList.sort_order))) or 0
-        row = ShoppingList(name=clean, sort_order=max_order + 1)
-        self.db.add(row)
-        self.db.commit()
-        self.db.refresh(row)
-        return {"ok": True, "list": self._list_to_dict(row, include_items=True), "created": True}
-
-    def rename_list(self, list_id: int, name: str) -> dict[str, Any]:
-        row = self._require_list(list_id)
-        clean = name.strip()
-        if not clean:
-            raise ValueError("Название списка не может быть пустым")
-        conflict = self.db.scalar(
-            select(ShoppingList).where(ShoppingList.name == clean, ShoppingList.id != list_id)
-        )
-        if conflict:
-            raise ValueError(f"Список «{clean}» уже существует")
-        row.name = clean
-        row.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        self.db.refresh(row)
-        return {"ok": True, "list": self._list_to_dict(row, include_items=True)}
-
-    def delete_list(self, list_id: int) -> dict[str, Any]:
-        row = self._require_list(list_id)
-        name = row.name
-        self.db.delete(row)
-        self.db.commit()
-        return {"ok": True, "deleted_list_id": list_id, "name": name}
-
-    def add_items(
-        self,
-        items: list[dict[str, Any]],
-        *,
-        list_id: int | None = None,
-        list_name: str | None = None,
-        create_list: bool = True,
-    ) -> dict[str, Any]:
-        if not items:
-            raise ValueError("Нужен хотя бы один товар")
-
-        row = self._resolve_list(list_id=list_id, name=list_name)
-        if not row and list_name and create_list:
-            created = self.create_list(list_name)
-            row = self._require_list(created["list"]["id"])
-        if not row:
-            raise ValueError("Укажи list_id или list_name")
-
-        max_order = self.db.scalar(
-            select(func.max(ShoppingListItem.sort_order)).where(ShoppingListItem.list_id == row.id)
-        ) or 0
-        added: list[dict[str, Any]] = []
-        for idx, raw in enumerate(items, start=1):
-            text = (raw.get("text") or "").strip()
-            if not text:
-                continue
-            item = ShoppingListItem(
-                list_id=row.id,
-                text=text,
-                quantity=raw.get("quantity"),
-                unit=(raw.get("unit") or "").strip(),
-                sort_order=max_order + idx,
-            )
-            self.db.add(item)
-            added.append(item)
-
-        if not added:
-            raise ValueError("Нет валидных товаров для добавления")
-
-        row.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        for item in added:
-            self.db.refresh(item)
-
-        return {
-            "ok": True,
-            "list_id": row.id,
-            "list_name": row.name,
-            "added": [self._item_to_dict(i) for i in added],
-            "list": self._list_to_dict(self._require_list(row.id), include_items=True),
-        }
-
-    def set_item_checked(self, item_id: int, checked: bool) -> dict[str, Any]:
-        item = self._require_item(item_id)
-        item.checked = checked
-        item.shopping_list.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        return {"ok": True, "item": self._item_to_dict(item)}
-
-    def remove_item(self, item_id: int) -> dict[str, Any]:
-        item = self._require_item(item_id)
-        data = self._item_to_dict(item)
-        list_id = item.list_id
-        self.db.delete(item)
-        self.db.commit()
-        return {"ok": True, "removed": data, "list_id": list_id}
-
-    def clear_checked(self, list_id: int) -> dict[str, Any]:
-        row = self._require_list(list_id)
-        removed = 0
-        for item in list(row.items):
-            if item.checked:
-                self.db.delete(item)
-                removed += 1
-        row.updated_at = datetime.now(timezone.utc)
-        self.db.commit()
-        return {
-            "ok": True,
-            "list_id": list_id,
-            "removed_count": removed,
-            "list": self._list_to_dict(self._require_list(list_id), include_items=True),
-        }
-
-    def _resolve_list(
-        self,
-        *,
-        list_id: int | None = None,
-        name: str | None = None,
-    ) -> ShoppingList | None:
-        if list_id is not None:
-            return self.db.scalar(
-                select(ShoppingList)
-                .where(ShoppingList.id == list_id)
-                .options(selectinload(ShoppingList.items))
-            )
-        if name:
-            clean = name.strip()
-            return self.db.scalar(
-                select(ShoppingList)
-                .where(ShoppingList.name == clean)
-                .options(selectinload(ShoppingList.items))
-            )
-        return None
-
-    def _require_list(self, list_id: int) -> ShoppingList:
-        row = self._resolve_list(list_id=list_id)
-        if not row:
-            raise ValueError(f"Список #{list_id} не найден")
-        return row
-
-    def _require_item(self, item_id: int) -> ShoppingListItem:
-        item = self.db.get(ShoppingListItem, item_id)
-        if not item:
-            raise ValueError(f"Позиция #{item_id} не найдена")
-        return item
-
-    def _item_to_dict(self, item: ShoppingListItem) -> dict[str, Any]:
-        return {
-            "id": item.id,
-            "list_id": item.list_id,
-            "text": item.text,
-            "quantity": item.quantity,
-            "unit": item.unit,
-            "checked": item.checked,
-            "sort_order": item.sort_order,
-        }
-
-    def _list_to_dict(self, row: ShoppingList, *, include_items: bool) -> dict[str, Any]:
-        data: dict[str, Any] = {
-            "id": row.id,
-            "name": row.name,
-            "sort_order": row.sort_order,
-            "item_count": len(row.items) if row.items is not None else 0,
-            "unchecked_count": sum(1 for i in (row.items or []) if not i.checked),
-        }
-        if include_items:
-            data["items"] = [self._item_to_dict(i) for i in (row.items or [])]
-        return data
+from datetime import datetime, timezone
+from typing import Any
+
+from sqlalchemy import func, select
+from sqlalchemy.orm import Session, selectinload
+
+from app.db.models import ShoppingList, ShoppingListItem
+
+
+class ShoppingService:
+    def __init__(self, db: Session, user_id: int):
+        self.db = db
+        self.user_id = user_id
+
+    def snapshot(self) -> dict[str, Any]:
+        lists = self.list_lists(include_items=True)
+        total_items = sum(len(lst.get("items") or []) for lst in lists)
+        unchecked = sum(
+            1
+            for lst in lists
+            for item in (lst.get("items") or [])
+            if not item.get("checked")
+        )
+        return {
+            "lists": lists,
+            "list_count": len(lists),
+            "total_items": total_items,
+            "unchecked_items": unchecked,
+        }
+
+    def list_lists(self, *, include_items: bool = False) -> list[dict[str, Any]]:
+        stmt = select(ShoppingList).where(ShoppingList.user_id == self.user_id).order_by(ShoppingList.sort_order, ShoppingList.name)
+        if include_items:
+            stmt = stmt.options(selectinload(ShoppingList.items))
+        rows = list(self.db.scalars(stmt).all())
+        return [self._list_to_dict(row, include_items=include_items) for row in rows]
+
+    def get_list(
+        self,
+        list_id: int | None = None,
+        *,
+        name: str | None = None,
+    ) -> dict[str, Any] | None:
+        row = self._resolve_list(list_id=list_id, name=name)
+        if not row:
+            return None
+        return self._list_to_dict(row, include_items=True)
+
+    def create_list(self, name: str) -> dict[str, Any]:
+        clean = name.strip()
+        if not clean:
+            raise ValueError("Название списка не может быть пустым")
+        existing = self.db.scalar(select(ShoppingList).where(ShoppingList.user_id == self.user_id, ShoppingList.name == clean))
+        if existing:
+            return {"ok": True, "list": self._list_to_dict(existing, include_items=True), "created": False}
+
+        max_order = self.db.scalar(select(func.max(ShoppingList.sort_order)).where(ShoppingList.user_id == self.user_id)) or 0
+        row = ShoppingList(user_id=self.user_id, name=clean, sort_order=max_order + 1)
+        self.db.add(row)
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "list": self._list_to_dict(row, include_items=True), "created": True}
+
+    def rename_list(self, list_id: int, name: str) -> dict[str, Any]:
+        row = self._require_list(list_id)
+        clean = name.strip()
+        if not clean:
+            raise ValueError("Название списка не может быть пустым")
+        conflict = self.db.scalar(
+            select(ShoppingList).where(ShoppingList.user_id == self.user_id, ShoppingList.name == clean, ShoppingList.id != list_id)
+        )
+        if conflict:
+            raise ValueError(f"Список «{clean}» уже существует")
+        row.name = clean
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        self.db.refresh(row)
+        return {"ok": True, "list": self._list_to_dict(row, include_items=True)}
+
+    def delete_list(self, list_id: int) -> dict[str, Any]:
+        row = self._require_list(list_id)
+        name = row.name
+        self.db.delete(row)
+        self.db.commit()
+        return {"ok": True, "deleted_list_id": list_id, "name": name}
+
+    def add_items(
+        self,
+        items: list[dict[str, Any]],
+        *,
+        list_id: int | None = None,
+        list_name: str | None = None,
+        create_list: bool = True,
+    ) -> dict[str, Any]:
+        if not items:
+            raise ValueError("Нужен хотя бы один товар")
+
+        row = self._resolve_list(list_id=list_id, name=list_name)
+        if not row and list_name and create_list:
+            created = self.create_list(list_name)
+            row = self._require_list(created["list"]["id"])
+        if not row:
+            raise ValueError("Укажи list_id или list_name")
+
+        max_order = self.db.scalar(
+            select(func.max(ShoppingListItem.sort_order)).where(ShoppingListItem.list_id == row.id)
+        ) or 0
+        added: list[dict[str, Any]] = []
+        for idx, raw in enumerate(items, start=1):
+            text = (raw.get("text") or "").strip()
+            if not text:
+                continue
+            item = ShoppingListItem(
+                list_id=row.id,
+                text=text,
+                quantity=raw.get("quantity"),
+                unit=(raw.get("unit") or "").strip(),
+                sort_order=max_order + idx,
+            )
+            self.db.add(item)
+            added.append(item)
+
+        if not added:
+            raise ValueError("Нет валидных товаров для добавления")
+
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        for item in added:
+            self.db.refresh(item)
+
+        return {
+            "ok": True,
+            "list_id": row.id,
+            "list_name": row.name,
+            "added": [self._item_to_dict(i) for i in added],
+            "list": self._list_to_dict(self._require_list(row.id), include_items=True),
+        }
+
+    def set_item_checked(self, item_id: int, checked: bool) -> dict[str, Any]:
+        item = self._require_item(item_id)
+        item.checked = checked
+        item.shopping_list.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return {"ok": True, "item": self._item_to_dict(item)}
+
+    def remove_item(self, item_id: int) -> dict[str, Any]:
+        item = self._require_item(item_id)
+        data = self._item_to_dict(item)
+        list_id = item.list_id
+        self.db.delete(item)
+        self.db.commit()
+        return {"ok": True, "removed": data, "list_id": list_id}
+
+    def clear_checked(self, list_id: int) -> dict[str, Any]:
+        row = self._require_list(list_id)
+        removed = 0
+        for item in list(row.items):
+            if item.checked:
+                self.db.delete(item)
+                removed += 1
+        row.updated_at = datetime.now(timezone.utc)
+        self.db.commit()
+        return {
+            "ok": True,
+            "list_id": list_id,
+            "removed_count": removed,
+            "list": self._list_to_dict(self._require_list(list_id), include_items=True),
+        }
+
+    def _resolve_list(
+        self,
+        *,
+        list_id: int | None = None,
+        name: str | None = None,
+    ) -> ShoppingList | None:
+        if list_id is not None:
+            return self.db.scalar(
+                select(ShoppingList)
+                .where(ShoppingList.user_id == self.user_id, ShoppingList.id == list_id)
+                .options(selectinload(ShoppingList.items))
+            )
+        if name:
+            clean = name.strip()
+            return self.db.scalar(
+                select(ShoppingList)
+                .where(ShoppingList.user_id == self.user_id, ShoppingList.name == clean)
+                .options(selectinload(ShoppingList.items))
+            )
+        return None
+
+    def _require_list(self, list_id: int) -> ShoppingList:
+        row = self._resolve_list(list_id=list_id)
+        if not row:
+            raise ValueError(f"Список #{list_id} не найден")
+        return row
+
+    def _require_item(self, item_id: int) -> ShoppingListItem:
+        item = self.db.get(ShoppingListItem, item_id)
+        if not item or item.shopping_list.user_id != self.user_id:
+            raise ValueError(f"Позиция #{item_id} не найдена")
+        return item
+
+    def _item_to_dict(self, item: ShoppingListItem) -> dict[str, Any]:
+        return {
+            "id": item.id,
+            "list_id": item.list_id,
+            "text": item.text,
+            "quantity": item.quantity,
+            "unit": item.unit,
+            "checked": item.checked,
+            "sort_order": item.sort_order,
+        }
+
+    def _list_to_dict(self, row: ShoppingList, *, include_items: bool) -> dict[str, Any]:
+        data: dict[str, Any] = {
+            "id": row.id,
+            "name": row.name,
+            "sort_order": row.sort_order,
+            "item_count": len(row.items) if row.items is not None else 0,
+            "unchecked_count": sum(1 for i in (row.items or []) if not i.checked),
+        }
+        if include_items:
+            data["items"] = [self._item_to_dict(i) for i in (row.items or [])]
+        return data

backend/app/tools/registry.py.refactor_bak

@@ -0,0 +1,961 @@
+import json
+from datetime import date, datetime, timedelta, timezone
+from typing import Any
+
+from sqlalchemy.orm import Session
+
+from app.fitness.service import FitnessService
+from app.fitness.structuring import structure_meal, structure_workout
+from app.homelab.digest import build_weather_briefing
+from app.homelab.image_gen import generate_image as run_generate_image
+from app.homelab.openmeteo import OpenMeteoClient
+from app.integrations.openfoodfacts import OpenFoodFactsClient
+from app.integrations.wger import WgerClient
+from app.memory.service import MemoryService
+from app.pomodoro.service import PomodoroService
+from app.projects.service import ProjectService
+from app.reminders.service import RemindersService
+from app.shopping.service import ShoppingService
+
+TOOL_DEFINITIONS: list[dict[str, Any]] = [
+    {
+        "type": "function",
+        "function": {
+            "name": "get_pomodoro_status",
+            "description": "ОБЯЗАТЕЛЬНО вызывай перед любым ответом о таймере. Статус, фаза и прогресс цикла.",
+            "parameters": {"type": "object", "properties": {}, "required": []},
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "start_pomodoro",
+            "description": "Запустить фазу работы в цикле помидоро (25 мин по умолчанию).",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "duration_min": {"type": "integer", "description": "Минуты работы"},
+                    "task_note": {"type": "string", "description": "Над чем работаем"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "start_short_break",
+            "description": "Запустить короткий перерыв между работами.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "duration_min": {"type": "integer", "description": "Минуты перерыва"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "start_long_break",
+            "description": "Запустить длинный перерыв после завершения цикла работ.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "duration_min": {"type": "integer", "description": "Минуты перерыва"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "stop_pomodoro",
+            "description": "Остановить текущую фазу таймера.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "result": {"type": "string", "description": "Отчёт о сделанном"},
+                    "completed": {
+                        "type": "boolean",
+                        "description": "True если фаза полностью завершена",
+                    },
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "skip_pomodoro_phase",
+            "description": "Досрочно завершить текущую фазу и перейти к следующей в цикле.",
+            "parameters": {"type": "object", "properties": {}, "required": []},
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "reset_pomodoro_cycle",
+            "description": "Сбросить цикл помидоро: обнулить счётчик работ и остановить таймер.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "clear_task": {
+                        "type": "boolean",
+                        "description": "Также очистить текущую задачу",
+                    },
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "get_pomodoro_history",
+            "description": "История помидоро-сессий (таймер), не Taiga-задачи.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "limit": {"type": "integer", "description": "Сколько сессий вернуть"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "sync_taiga_projects",
+            "description": "Синхронизировать список проектов из Taiga API. Вызывай если проекты неизвестны.",
+            "parameters": {"type": "object", "properties": {}, "required": []},
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "list_taiga_projects",
+            "description": "Список проектов Taiga с привязкой Gitea.",
+            "parameters": {"type": "object", "properties": {}, "required": []},
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "list_taiga_tasks",
+            "description": (
+                "ОБЯЗАТЕЛЬНО при вопросах «какие задачи», «покажи задачи проекта», «что открыто в Taiga». "
+                "Живые user stories и tasks из Taiga API. НЕ путать с list_work_items."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "project_slug": {
+                        "type": "string",
+                        "description": "Slug проекта, например home_assistant. Пусто = все проекты.",
+                    },
+                    "limit": {"type": "integer", "description": "Макс. на проект, по умолчанию 20"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "create_work_item",
+            "description": (
+                "Создать фичу/баг из вольного текста: структурировать через LLM, "
+                "создать Taiga story + Gitea issue. Вызывай при «заведи баг», «оформи фичу», «добавь в таигу»."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "text": {"type": "string", "description": "Полное описание от пользователя"},
+                    "project_slug": {
+                        "type": "string",
+                        "description": "Slug проекта Taiga, если известен",
+                    },
+                },
+                "required": ["text"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "remember_fact",
+            "description": (
+                "Сохранить факт в долгосрочную память. "
+                "Когда пользователь просит «запомни», или сообщает устойчивое предпочтение/факт."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "content": {"type": "string", "description": "Что запомнить"},
+                    "category": {
+                        "type": "string",
+                        "description": "preference, person, habit, project, fact",
+                    },
+                    "importance": {"type": "integer", "description": "1-5, по умолчанию 3"},
+                },
+                "required": ["content"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "recall_memories",
+            "description": (
+                "Поиск в долгосрочной памяти. "
+                "Когда спрашивают «что ты помнишь», «что я говорил про X»."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "query": {"type": "string", "description": "Подстрока для поиска"},
+                    "category": {"type": "string"},
+                    "limit": {"type": "integer"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "forget_memory",
+            "description": "Удалить (деактивировать) факт по id из recall_memories или снимка памяти.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "memory_id": {"type": "integer"},
+                },
+                "required": ["memory_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "update_profile",
+            "description": (
+                "Обновить профиль пользователя: name, timezone, language, notes. "
+                "Передавай только изменившиеся поля."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "name": {"type": "string"},
+                    "age": {"type": "string", "description": "Возраст пользователя"},
+                    "timezone": {"type": "string"},
+                    "language": {"type": "string"},
+                    "notes": {"type": "string"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "update_session_summary",
+            "description": (
+                "Сохранить краткую сводку темы текущего чата "
+                "(когда диалог длинный или пользователь просит «сожми контекст»)."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "summary": {"type": "string", "description": "2-5 предложений о теме чата"},
+                    "session_id": {"type": "integer"},
+                },
+                "required": ["summary", "session_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "get_fitness_summary",
+            "description": (
+                "Сводка фитнеса за день: ккал, БЖУ, вода, еда, тренировки. "
+                "Без даты — сегодня; date=YYYY-MM-DD или days_ago=1 (вчера)."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "date": {"type": "string", "description": "Дата YYYY-MM-DD"},
+                    "days_ago": {
+                        "type": "integer",
+                        "description": "0 сегодня, 1 вчера, 2 позавчера…",
+                    },
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "get_fitness_history",
+            "description": (
+                "Краткая история за несколько дней (ккал, вода, тренировки по дням). "
+                "«На прошлой неделе», «за 7 дней»."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "days": {"type": "integer", "description": "Сколько дней, по умолчанию 7"},
+                    "end_date": {"type": "string", "description": "Конец периода YYYY-MM-DD, по умолчанию сегодня"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "set_fitness_profile",
+            "description": "Настроить фитнес-профиль и пересчитать цели ккал/БЖУ/воды.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "sex": {"type": "string", "description": "male/female"},
+                    "age": {"type": "integer"},
+                    "height_cm": {"type": "number"},
+                    "weight_kg": {"type": "number"},
+                    "activity_level": {
+                        "type": "string",
+                        "description": "sedentary/light/moderate/active/very_active",
+                    },
+                    "goal": {"type": "string", "description": "lose/maintain/gain"},
+                    "target_weight_kg": {"type": "number"},
+                    "weekly_workouts": {"type": "integer"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "calc_fitness_targets",
+            "description": "Калькулятор BMR/TDEE/макросов без сохранения.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "sex": {"type": "string"},
+                    "age": {"type": "integer"},
+                    "height_cm": {"type": "number"},
+                    "weight_kg": {"type": "number"},
+                    "activity_level": {"type": "string"},
+                    "goal": {"type": "string"},
+                },
+                "required": ["weight_kg", "height_cm", "age"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "log_meal",
+            "description": "Записать приём пищи. LLM оценит ккал и БЖУ из текста.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "text": {"type": "string", "description": "Что съел"},
+                    "meal_type": {"type": "string"},
+                },
+                "required": ["text"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "log_water",
+            "description": "Записать воду в мл.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "amount_ml": {"type": "integer"},
+                },
+                "required": ["amount_ml"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "log_weight",
+            "description": "Записать вес в кг.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "weight_kg": {"type": "number"},
+                    "body_fat_pct": {"type": "number"},
+                    "notes": {"type": "string"},
+                },
+                "required": ["weight_kg"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "log_workout",
+            "description": "Записать тренировку из текста.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "text": {"type": "string"},
+                },
+                "required": ["text"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "lookup_food",
+            "description": "Поиск продукта в Open Food Facts (ккал на 100г).",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "query": {"type": "string"},
+                    "limit": {"type": "integer"},
+                },
+                "required": ["query"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "lookup_exercise",
+            "description": "Поиск упражнения в базе wger.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "query": {"type": "string"},
+                    "limit": {"type": "integer"},
+                },
+                "required": ["query"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "set_fitness_reminder",
+            "description": "Вкл/выкл или настроить напоминание: water, meal, workout, weigh_in.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "kind": {"type": "string"},
+                    "enabled": {"type": "boolean"},
+                    "hour": {"type": "integer"},
+                    "minute": {"type": "integer"},
+                    "interval_hours": {"type": "integer"},
+                },
+                "required": ["kind"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "get_weather",
+            "description": (
+                "ОБЯЗАТЕЛЬНО для вопросов о погоде, «что на улице», «будет ли дождь». "
+                "Текущая погода и прогноз по часам."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "hours_ahead": {
+                        "type": "integer",
+                        "description": "Сколько часов прогноза (по умолчанию 12)",
+                    },
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "get_morning_briefing",
+            "description": "Утренний брифинг: погода и заголовки новостей.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "include_news": {
+                        "type": "boolean",
+                        "description": "Включить новости (по умолчанию true)",
+                    },
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "generate_image",
+            "description": (
+                "Аниме-картинка (Anima через RP-чат). "
+                "«Нарисуй себя» / портрет персонажа → draw_self=true. "
+                "Другая сцена → scene_description на английском (booru-теги). "
+                "Внешность берётся из карточки персонажа. Только по запросу или когда уместно."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "draw_self": {
+                        "type": "boolean",
+                        "description": "Нарисовать персонажа из карточки в контексте текущего чата",
+                    },
+                    "scene_description": {
+                        "type": "string",
+                        "description": "Описание сцены на английском (booru-теги), если не draw_self",
+                    },
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "list_shopping_lists",
+            "description": "Все списки покупок с позициями. «Что купить», «покажи списки».",
+            "parameters": {"type": "object", "properties": {}, "required": []},
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "create_shopping_list",
+            "description": "Создать новый список покупок.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "name": {"type": "string", "description": "Название списка, например «Продукты»"},
+                },
+                "required": ["name"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "add_shopping_items",
+            "description": "Добавить товары в список. Список создаётся, если не существует.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "list_name": {"type": "string", "description": "Название списка"},
+                    "list_id": {"type": "integer"},
+                    "items": {
+                        "type": "array",
+                        "items": {
+                            "type": "object",
+                            "properties": {
+                                "text": {"type": "string"},
+                                "quantity": {"type": "number"},
+                                "unit": {"type": "string"},
+                            },
+                            "required": ["text"],
+                        },
+                    },
+                },
+                "required": ["items"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "check_shopping_item",
+            "description": "Отметить позицию как купленную (checked=true) или снять отметку (false).",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "item_id": {"type": "integer"},
+                    "checked": {"type": "boolean"},
+                },
+                "required": ["item_id", "checked"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "remove_shopping_item",
+            "description": "Удалить позицию из списка по item_id.",
+            "parameters": {
+                "type": "object",
+                "properties": {"item_id": {"type": "integer"}},
+                "required": ["item_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "delete_shopping_list",
+            "description": "Удалить весь список покупок.",
+            "parameters": {
+                "type": "object",
+                "properties": {"list_id": {"type": "integer"}},
+                "required": ["list_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "list_reminders",
+            "description": "Список активных напоминаний. «Что напомнил», «мои напоминания».",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "limit": {"type": "integer", "description": "Макс. записей, по умолчанию 20"},
+                },
+                "required": [],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "create_reminder",
+            "description": (
+                "Создать напоминание. due_at — ISO datetime в часовом поясе пользователя "
+                "(см. [Текущее время]). Примеры: через 15 мин, завтра 09:00, 2027-05-12T12:16:00."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "title": {"type": "string", "description": "О чём напомнить"},
+                    "due_at": {"type": "string", "description": "ISO datetime"},
+                    "notes": {"type": "string"},
+                    "all_day": {"type": "boolean"},
+                    "recurrence": {
+                        "type": "string",
+                        "enum": ["none", "daily", "weekly", "monthly", "yearly"],
+                        "description": "Повтор (yearly — день рождения, Новый год)",
+                    },
+                },
+                "required": ["title", "due_at"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "update_reminder",
+            "description": "Изменить напоминание по id.",
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "reminder_id": {"type": "integer"},
+                    "title": {"type": "string"},
+                    "due_at": {"type": "string"},
+                    "notes": {"type": "string"},
+                    "all_day": {"type": "boolean"},
+                    "recurrence": {
+                        "type": "string",
+                        "enum": ["none", "daily", "weekly", "monthly", "yearly"],
+                    },
+                    "enabled": {"type": "boolean"},
+                },
+                "required": ["reminder_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "delete_reminder",
+            "description": "Удалить напоминание по id.",
+            "parameters": {
+                "type": "object",
+                "properties": {"reminder_id": {"type": "integer"}},
+                "required": ["reminder_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "complete_reminder",
+            "description": "Отметить напоминание выполненным (снять с календаря).",
+            "parameters": {
+                "type": "object",
+                "properties": {"reminder_id": {"type": "integer"}},
+                "required": ["reminder_id"],
+            },
+        },
+    },
+    {
+        "type": "function",
+        "function": {
+            "name": "list_work_items",
+            "description": (
+                "Только задачи, созданные ЭТИМ ассистентом через create_work_item (локальная БД). "
+                "НЕ использовать для общего вопроса «какие задачи в Taiga» — для того list_taiga_tasks."
+            ),
+            "parameters": {
+                "type": "object",
+                "properties": {
+                    "status": {"type": "string", "description": "open или closed"},
+                    "limit": {"type": "integer"},
+                },
+                "required": [],
+            },
+        },
+    },
+]
+
+
+async def execute_tool(
+    db: Session,
+    name: str,
+    arguments: dict[str, Any],
+    *,
+    session_id: int | None = None,
+) -> str:
+    pomodoro = PomodoroService(db)
+    projects = ProjectService(db)
+    memory = MemoryService(db)
+    fitness = FitnessService(db)
+    shopping = ShoppingService(db)
+    reminders = RemindersService(db)
+
+    try:
+        if name == "get_pomodoro_status":
+            result = pomodoro.get_status()
+        elif name == "start_pomodoro":
+            result = pomodoro.start_work(
+                duration_min=arguments.get("duration_min"),
+                task_note=arguments.get("task_note", ""),
+            )
+        elif name == "start_short_break":
+            result = pomodoro.start_short_break(duration_min=arguments.get("duration_min"))
+        elif name == "start_long_break":
+            result = pomodoro.start_long_break(duration_min=arguments.get("duration_min"))
+        elif name == "stop_pomodoro":
+            result = pomodoro.stop(
+                result=arguments.get("result", ""),
+                completed=arguments.get("completed", False),
+            )
+        elif name == "skip_pomodoro_phase":
+            result = pomodoro.skip_phase()
+        elif name == "reset_pomodoro_cycle":
+            result = pomodoro.reset_cycle(clear_task=arguments.get("clear_task", False))
+        elif name == "get_pomodoro_history":
+            result = pomodoro.history(limit=arguments.get("limit", 10))
+        elif name == "sync_taiga_projects":
+            from app.projects.context import invalidate_projects_snapshot_cache
+
+            result = projects.sync_taiga_projects()
+            invalidate_projects_snapshot_cache()
+        elif name == "list_taiga_projects":
+            result = projects.list_projects()
+        elif name == "list_taiga_tasks":
+            result = projects.list_taiga_open_tasks(
+                project_slug=arguments.get("project_slug"),
+                limit=arguments.get("limit", 20),
+            )
+        elif name == "create_work_item":
+            result = await projects.create_work_item(
+                arguments.get("text", ""),
+                project_slug=arguments.get("project_slug"),
+            )
+        elif name == "list_work_items":
+            result = projects.list_work_items(
+                limit=arguments.get("limit", 20),
+                status=arguments.get("status"),
+            )
+        elif name == "remember_fact":
+            result = memory.remember_fact(
+                arguments.get("content", ""),
+                category=arguments.get("category", "fact"),
+                importance=arguments.get("importance", 3),
+                session_id=session_id,
+                source="tool",
+            )
+        elif name == "recall_memories":
+            result = memory.recall_memories(
+                query=arguments.get("query"),
+                category=arguments.get("category"),
+                limit=arguments.get("limit", 20),
+            )
+        elif name == "forget_memory":
+            result = memory.forget_memory(int(arguments["memory_id"]))
+        elif name == "update_profile":
+            updates = {
+                k: arguments[k]
+                for k in ("name", "age", "timezone", "language", "notes")
+                if k in arguments and arguments[k] is not None
+            }
+            result = memory.update_profile(updates)
+        elif name == "update_session_summary":
+            result = memory.update_session_summary(
+                int(arguments["session_id"]),
+                arguments.get("summary", ""),
+            )
+        elif name == "get_fitness_summary":
+            day: date | None = None
+            if arguments.get("date"):
+                day = date.fromisoformat(str(arguments["date"]))
+            elif arguments.get("days_ago") is not None:
+                day = datetime.now(timezone.utc).date() - timedelta(
+                    days=int(arguments["days_ago"])
+                )
+            result = fitness.get_daily_summary(day)
+        elif name == "get_fitness_history":
+            end_day = None
+            if arguments.get("end_date"):
+                end_day = date.fromisoformat(str(arguments["end_date"]))
+            result = fitness.get_history(
+                days=int(arguments.get("days") or 7),
+                end_day=end_day,
+            )
+        elif name == "set_fitness_profile":
+            updates = {
+                k: arguments[k]
+                for k in (
+                    "sex", "age", "height_cm", "weight_kg", "activity_level",
+                    "goal", "target_weight_kg", "weekly_workouts",
+                )
+                if k in arguments and arguments[k] is not None
+            }
+            result = fitness.set_profile(updates)
+        elif name == "calc_fitness_targets":
+            result = fitness.calc_targets(arguments)
+        elif name == "log_meal":
+            structured = await structure_meal(arguments.get("text", ""))
+            result = fitness.log_meal(
+                description=structured.get("description") or arguments.get("text", ""),
+                meal_type=arguments.get("meal_type") or structured.get("meal_type") or "snack",
+                calories=float(structured.get("calories") or 0),
+                protein_g=float(structured.get("protein_g") or 0),
+                fat_g=float(structured.get("fat_g") or 0),
+                carbs_g=float(structured.get("carbs_g") or 0),
+                source="llm",
+                estimated=True,
+            )
+        elif name == "log_water":
+            result = fitness.log_water(int(arguments.get("amount_ml", 250)))
+        elif name == "log_weight":
+            result = fitness.log_weight(
+                float(arguments["weight_kg"]),
+                body_fat_pct=arguments.get("body_fat_pct"),
+                notes=arguments.get("notes", ""),
+            )
+        elif name == "log_workout":
+            structured = await structure_workout(arguments.get("text", ""))
+            result = fitness.log_workout(
+                title=structured.get("title") or "Тренировка",
+                notes=structured.get("notes") or arguments.get("text", ""),
+                duration_min=structured.get("duration_min"),
+                exercises=structured.get("exercises"),
+            )
+        elif name == "lookup_food":
+            result = OpenFoodFactsClient().search(
+                arguments.get("query", ""),
+                limit=arguments.get("limit", 5),
+            )
+        elif name == "lookup_exercise":
+            result = WgerClient().search_exercises(
+                arguments.get("query", ""),
+                limit=arguments.get("limit", 8),
+            )
+        elif name == "set_fitness_reminder":
+            result = fitness.set_reminder(
+                arguments.get("kind", "water"),
+                enabled=arguments.get("enabled"),
+                hour=arguments.get("hour"),
+                minute=arguments.get("minute"),
+                interval_hours=arguments.get("interval_hours"),
+            )
+        elif name == "get_weather":
+            hours = int(arguments.get("hours_ahead") or 12)
+            client = OpenMeteoClient()
+            weather = client.fetch_current_and_hourly(hours_ahead=hours)
+            result = {
+                "weather": weather,
+                "rain_summary": client.rain_summary(hours_ahead=hours) if weather.get("ok") else "",
+            }
+        elif name == "get_morning_briefing":
+            include_news = arguments.get("include_news", True)
+            result = build_weather_briefing(
+                hours_ahead=12,
+                include_news=bool(include_news),
+            )
+        elif name == "generate_image":
+            result = await run_generate_image(
+                db,
+                session_id=session_id,
+                draw_self=bool(arguments.get("draw_self")),
+                scene_description=arguments.get("scene_description", ""),
+            )
+        elif name == "list_shopping_lists":
+            result = shopping.list_lists(include_items=True)
+        elif name == "create_shopping_list":
+            result = shopping.create_list(arguments.get("name", ""))
+        elif name == "add_shopping_items":
+            result = shopping.add_items(
+                arguments.get("items") or [],
+                list_id=arguments.get("list_id"),
+                list_name=arguments.get("list_name"),
+            )
+        elif name == "check_shopping_item":
+            result = shopping.set_item_checked(
+                int(arguments["item_id"]),
+                bool(arguments.get("checked", True)),
+            )
+        elif name == "remove_shopping_item":
+            result = shopping.remove_item(int(arguments["item_id"]))
+        elif name == "delete_shopping_list":
+            result = shopping.delete_list(int(arguments["list_id"]))
+        elif name == "list_reminders":
+            result = reminders.list_upcoming(limit=int(arguments.get("limit") or 20))
+        elif name == "create_reminder":
+            result = reminders.create(
+                title=arguments.get("title", ""),
+                due_at=arguments.get("due_at", ""),
+                notes=arguments.get("notes", ""),
+                all_day=bool(arguments.get("all_day", False)),
+                recurrence=arguments.get("recurrence", "none"),
+            )
+        elif name == "update_reminder":
+            result = reminders.update(
+                int(arguments["reminder_id"]),
+                title=arguments.get("title"),
+                due_at=arguments.get("due_at"),
+                notes=arguments.get("notes"),
+                all_day=arguments.get("all_day"),
+                recurrence=arguments.get("recurrence"),
+                enabled=arguments.get("enabled"),
+            )
+        elif name == "delete_reminder":
+            result = reminders.delete(int(arguments["reminder_id"]))
+        elif name == "complete_reminder":
+            result = reminders.complete(int(arguments["reminder_id"]))
+        else:
+            return json.dumps({"error": f"Unknown tool: {name}"}, ensure_ascii=False)
+
+        return json.dumps(result, ensure_ascii=False)
+    except ValueError as exc:
+        return json.dumps({"error": str(exc)}, ensure_ascii=False)
+    except Exception as exc:
+        return json.dumps({"error": str(exc)}, ensure_ascii=False)

backend/backfill_fitness_activity.py

@@ -0,0 +1,72 @@
+"""Backfill workout active_calories / steps from notes via regex."""
+from __future__ import annotations
+
+import re
+import sys
+from pathlib import Path
+
+sys.path.insert(0, str(Path(__file__).resolve().parents[1]))
+
+from sqlalchemy import select
+
+from app.db.base import SessionLocal
+from app.db.models import WorkoutLog
+
+ACTIVE_PATTERNS = [
+    re.compile(r"(?:active|активн(?:ые|ых)?)\s*(?:калор|kcal|ккал)[^\d]*(\d+(?:[.,]\d+)?)", re.I),
+    re.compile(r"(\d+(?:[.,]\d+)?)\s*(?:ккал|kcal)\s*(?:active|актив)", re.I),
+    re.compile(r"актив[^\d]{0,20}(\d+(?:[.,]\d+)?)", re.I),
+]
+TOTAL_PATTERNS = [
+    re.compile(r"(?:total|всего|сожжено|burned)[^\d]*(\d+(?:[.,]\d+)?)\s*(?:ккал|kcal)", re.I),
+    re.compile(r"(\d+(?:[.,]\d+)?)\s*(?:ккал|kcal)\s*(?:total|всего)", re.I),
+]
+STEPS_PATTERNS = [
+    re.compile(r"(?:шаг|step)s?\s*[:\-]?\s*(\d+)", re.I),
+    re.compile(r"(\d+)\s*(?:шаг|steps)", re.I),
+]
+
+
+def _first(patterns: list[re.Pattern[str]], text: str, *, as_int: bool = False):
+    for pat in patterns:
+        m = pat.search(text)
+        if not m:
+            continue
+        raw = m.group(1).replace(",", ".")
+        return int(float(raw)) if as_int else float(raw)
+    return None
+
+
+def main() -> None:
+    db = SessionLocal()
+    updated = 0
+    try:
+        rows = db.scalars(select(WorkoutLog)).all()
+        for row in rows:
+            text = f"{row.title or ''}\n{row.notes or ''}"
+            changed = False
+            if row.active_calories is None:
+                val = _first(ACTIVE_PATTERNS, text)
+                if val is not None:
+                    row.active_calories = float(val)
+                    changed = True
+            if row.total_calories is None:
+                val = _first(TOTAL_PATTERNS, text)
+                if val is not None:
+                    row.total_calories = float(val)
+                    changed = True
+            if row.steps is None:
+                val = _first(STEPS_PATTERNS, text, as_int=True)
+                if val is not None:
+                    row.steps = int(val)
+                    changed = True
+            if changed:
+                updated += 1
+        db.commit()
+        print(f"updated {updated} workout rows")
+    finally:
+        db.close()
+
+
+if __name__ == "__main__":
+    main()

backend/requirements.txt

@@ -1,9 +1,11 @@
-fastapi>=0.115.0
-uvicorn[standard]>=0.32.0
-sqlalchemy>=2.0.36
-pydantic-settings>=2.6.0
-openai>=1.55.0
-python-dotenv>=1.0.1
-aiosqlite>=0.20.0
-httpx>=0.28.0
-feedparser>=6.0.11
+fastapi>=0.115.0
+uvicorn[standard]>=0.32.0
+python-multipart>=0.0.9
+sqlalchemy>=2.0.36
+pydantic-settings>=2.6.0
+openai>=1.55.0
+python-dotenv>=1.0.1
+aiosqlite>=0.20.0
+httpx>=0.28.0
+feedparser>=6.0.11
+qdrant-client>=1.12.0,<1.13.0

backend/requirements.txt.refactor_bak

@@ -0,0 +1,9 @@
+fastapi>=0.115.0
+uvicorn[standard]>=0.32.0
+sqlalchemy>=2.0.36
+pydantic-settings>=2.6.0
+openai>=1.55.0
+python-dotenv>=1.0.1
+aiosqlite>=0.20.0
+httpx>=0.28.0
+feedparser>=6.0.11

backend/scripts/create_user.py

@@ -0,0 +1,42 @@
+"""Create a user with API token. Usage:
+  python -m scripts.create_user testuser --display-name "Test User"
+  python -m scripts.create_user guest --token my-custom-token-32chars-min
+"""
+from __future__ import annotations
+
+import argparse
+import sys
+
+from app.db.base import SessionLocal, init_db
+from app.auth.service import create_user
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser(description="Create Home Assistant user")
+    parser.add_argument("username", help="Unique username (lowercase)")
+    parser.add_argument("--display-name", default="", help="Display name")
+    parser.add_argument("--token", default="", help="Custom API token (auto-generated if empty)")
+    args = parser.parse_args()
+
+    init_db()
+    db = SessionLocal()
+    try:
+        user, plain_token = create_user(
+            db,
+            username=args.username,
+            display_name=args.display_name or args.username,
+            api_token=args.token or None,
+        )
+    except ValueError as exc:
+        print(f"Error: {exc}", file=sys.stderr)
+        return 1
+    finally:
+        db.close()
+
+    print(f"Created user id={user.id} username={user.username}")
+    print(f"API token (save it, shown once):\n{plain_token}")
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

backend/tests/test_activity_budget.py

@@ -0,0 +1,84 @@
+import unittest
+
+from app.fitness.activity_budget import (
+    compute_activity_bonus,
+    scale_targets,
+    steps_bonus_kcal,
+)
+
+
+PROFILE = {
+    "weight_kg": 70,
+    "activity_level": "moderate",
+    "weekly_workouts": 3,
+    "calorie_target": 2000,
+    "protein_g": 126,
+    "fat_g": 56,
+    "carbs_g": 250,
+    "water_l": 2.5,
+}
+
+
+class ActivityBudgetTests(unittest.TestCase):
+    def test_no_bonus_at_baseline(self) -> None:
+        bonus = compute_activity_bonus(
+            PROFILE,
+            steps_total=9000,
+            workouts=[{"active_calories": 85}],
+        )
+        self.assertEqual(bonus.steps_bonus_kcal, 0.0)
+        self.assertEqual(bonus.workout_bonus_kcal, 0.0)
+        self.assertEqual(bonus.total_bonus_kcal, 0.0)
+        self.assertEqual(bonus.scale_factor, 1.0)
+
+    def test_steps_and_workout_bonus(self) -> None:
+        bonus = compute_activity_bonus(
+            PROFILE,
+            steps_total=21800,
+            workouts=[{"active_calories": 155}],
+        )
+        self.assertGreater(bonus.steps_bonus_kcal, 0)
+        self.assertGreater(bonus.workout_bonus_kcal, 0)
+        self.assertEqual(
+            bonus.total_bonus_kcal,
+            round(bonus.steps_bonus_kcal + bonus.workout_bonus_kcal, 1),
+        )
+        self.assertGreater(bonus.scale_factor, 1.0)
+
+    def test_steps_bonus_formula(self) -> None:
+        kcal = steps_bonus_kcal(steps=21800, baseline_steps=9000, weight_kg=70)
+        self.assertEqual(kcal, round(12800 * 70 * 0.0005, 1))
+
+    def test_proportional_macro_scale(self) -> None:
+        base = {
+            "calories": 2000,
+            "protein_g": 100,
+            "fat_g": 50,
+            "carbs_g": 200,
+            "water_ml": 2500,
+        }
+        effective, targets_base = scale_targets(base, 500)
+        self.assertEqual(effective["calories"], 2500)
+        self.assertEqual(targets_base, base)
+        self.assertEqual(effective["water_ml"], 2500)
+        ratio = effective["calories"] / base["calories"]
+        self.assertAlmostEqual(effective["protein_g"] / base["protein_g"], ratio, places=1)
+        self.assertAlmostEqual(effective["fat_g"] / base["fat_g"], ratio, places=1)
+        self.assertAlmostEqual(effective["carbs_g"] / base["carbs_g"], ratio, places=1)
+
+    def test_floor_at_base_when_no_activity(self) -> None:
+        effective, _ = scale_targets(
+            {
+                "calories": 2045,
+                "protein_g": 156,
+                "fat_g": 57,
+                "carbs_g": 227,
+                "water_ml": 2900,
+            },
+            0,
+        )
+        self.assertEqual(effective["calories"], 2045)
+
+
+if __name__ == "__main__":
+    unittest.main()

backend/tests/test_body_composition.py

@@ -0,0 +1,102 @@
+import pytest
+
+from app.fitness.body_composition import (
+    compute_body_composition,
+    ffmi,
+    lean_body_mass,
+    navy_body_fat_pct,
+    whr,
+)
+
+
+def test_navy_male_reasonable_range():
+    bf = navy_body_fat_pct(
+        sex="male",
+        height_cm=180,
+        neck_cm=38,
+        waist_cm=84,
+    )
+    assert bf is not None
+    assert 5 <= bf <= 35
+
+
+def test_navy_female_requires_hip():
+    assert navy_body_fat_pct(
+        sex="female",
+        height_cm=165,
+        neck_cm=34,
+        waist_cm=72,
+        hip_cm=None,
+    ) is None
+
+    bf = navy_body_fat_pct(
+        sex="female",
+        height_cm=165,
+        neck_cm=34,
+        waist_cm=72,
+        hip_cm=98,
+    )
+    assert bf is not None
+    assert 10 <= bf <= 45
+
+
+def test_navy_invalid_waist_neck():
+    assert navy_body_fat_pct(
+        sex="male",
+        height_cm=180,
+        neck_cm=40,
+        waist_cm=39,
+    ) is None
+
+
+def test_whr():
+    assert whr(80, 100) == 0.8
+
+
+def test_lean_body_mass_and_ffmi():
+    lbm = lean_body_mass(80, 20)
+    assert lbm == 64.0
+    score = ffmi(80, 180, 20)
+    assert score is not None
+    assert 15 <= score <= 25
+
+
+def test_compute_manual_body_fat():
+    result = compute_body_composition(
+        sex="male",
+        height_cm=180,
+        weight_kg=80,
+        body_fat_pct=18,
+        waist_cm=84,
+        hip_cm=100,
+    )
+    assert result["body_fat_pct"] == 18.0
+    assert result["body_fat_method"] == "manual"
+    assert result["whr"] == 0.84
+    assert result["lbm_kg"] == 65.6
+    assert result["ffmi"] is not None
+
+
+def test_compute_navy_auto():
+    result = compute_body_composition(
+        sex="male",
+        height_cm=180,
+        weight_kg=82,
+        neck_cm=38,
+        waist_cm=84,
+    )
+    assert result["body_fat_pct"] is not None
+    assert result["body_fat_method"] == "navy"
+    assert result["lbm_kg"] is not None
+
+
+def test_compute_female_warning_without_hip():
+    result = compute_body_composition(
+        sex="female",
+        height_cm=165,
+        weight_kg=60,
+        neck_cm=34,
+        waist_cm=72,
+    )
+    assert result["body_fat_pct"] is None
+    assert any("бёдер" in w for w in result["warnings"])

backend/tests/test_multi_user.py

@@ -0,0 +1,154 @@
+import os
+import tempfile
+import uuid
+from pathlib import Path
+
+import pytest
+from fastapi.testclient import TestClient
+from sqlalchemy import create_engine
+from sqlalchemy.orm import sessionmaker
+from sqlalchemy.pool import StaticPool
+
+from app.auth.tokens import hash_token
+from app.db.base import Base, get_db
+from app.db.models import CharacterCard, ChatSession, MemoryFact, ShoppingList, User
+
+
+@pytest.fixture()
+def client():
+    db_path = Path(tempfile.gettempdir()) / f"test_multi_{uuid.uuid4().hex}.db"
+    os.environ["DATABASE_URL"] = f"sqlite:///{db_path.as_posix()}"
+    os.environ["DEFAULT_API_TOKEN"] = "unused-in-tests"
+    os.environ["AUTH_REQUIRED"] = "true"
+    os.environ["RAG_ENABLED"] = "false"
+
+    from app.config import get_settings
+
+    get_settings.cache_clear()
+
+    from app.main import create_app
+
+    engine = create_engine(
+        f"sqlite:///{db_path.as_posix()}",
+        connect_args={"check_same_thread": False},
+        poolclass=StaticPool,
+    )
+    TestingSessionLocal = sessionmaker(autocommit=False, autoflush=False, bind=engine)
+    Base.metadata.create_all(bind=engine)
+
+    token_a = "token-user-a"
+    token_b = "token-user-b"
+
+    db = TestingSessionLocal()
+    user_a = User(
+        username="alice",
+        display_name="Alice",
+        api_token_hash=hash_token(token_a),
+        is_active=True,
+    )
+    user_b = User(
+        username="bob",
+        display_name="Bob",
+        api_token_hash=hash_token(token_b),
+        is_active=True,
+    )
+    db.add_all([user_a, user_b])
+    db.commit()
+    db.refresh(user_a)
+    db.refresh(user_b)
+
+    db.add(ChatSession(user_id=user_a.id, title="Alice chat"))
+    db.add(ChatSession(user_id=user_b.id, title="Bob chat"))
+    db.add(ShoppingList(user_id=user_a.id, name="groceries"))
+    db.add(ShoppingList(user_id=user_b.id, name="groceries"))
+    db.add(
+        CharacterCard(
+            user_id=user_a.id,
+            card_json='{"spec":"chara_card_v2","spec_version":"2.0","data":{"name":"A","rp_persona_id":"persona-a"}}',
+        )
+    )
+    db.add(
+        CharacterCard(
+            user_id=user_b.id,
+            card_json='{"spec":"chara_card_v2","spec_version":"2.0","data":{"name":"B","rp_persona_id":"persona-b"}}',
+        )
+    )
+    db.add(
+        MemoryFact(
+            user_id=user_a.id,
+            category="person",
+            content="Секрет только для owner",
+            source="test",
+        )
+    )
+    db.commit()
+    db.close()
+
+    app = create_app()
+
+    def override_get_db():
+        db = TestingSessionLocal()
+        try:
+            yield db
+        finally:
+            db.close()
+
+    app.dependency_overrides[get_db] = override_get_db
+    with TestClient(app) as test_client:
+        test_client.tokens = {"a": token_a, "b": token_b}
+        yield test_client
+
+    app.dependency_overrides.clear()
+    get_settings.cache_clear()
+    try:
+        db_path.unlink(missing_ok=True)
+    except OSError:
+        pass
+
+
+def _headers(client: TestClient, who: str) -> dict[str, str]:
+    return {"Authorization": f"Bearer {client.tokens[who]}"}
+
+
+def test_chat_sessions_isolated(client: TestClient):
+    res_a = client.get("/api/v1/chat/sessions", headers=_headers(client, "a"))
+    res_b = client.get("/api/v1/chat/sessions", headers=_headers(client, "b"))
+    assert res_a.status_code == 200
+    assert res_b.status_code == 200
+    titles_a = {s["title"] for s in res_a.json()}
+    titles_b = {s["title"] for s in res_b.json()}
+    assert titles_a == {"Alice chat"}
+    assert titles_b == {"Bob chat"}
+
+
+def test_character_cards_isolated(client: TestClient):
+    res_a = client.get("/api/v1/character", headers=_headers(client, "a"))
+    res_b = client.get("/api/v1/character", headers=_headers(client, "b"))
+    assert res_a.json()["data"]["rp_persona_id"] == "persona-a"
+    assert res_b.json()["data"]["rp_persona_id"] == "persona-b"
+
+
+def test_shopping_same_name_different_users(client: TestClient):
+    res_a = client.get("/api/v1/shopping", headers=_headers(client, "a"))
+    res_b = client.get("/api/v1/shopping", headers=_headers(client, "b"))
+    assert res_a.status_code == 200
+    assert res_b.status_code == 200
+    assert len(res_a.json()["lists"]) == 1
+    assert len(res_b.json()["lists"]) == 1
+
+
+def test_missing_token_unauthorized(client: TestClient):
+    res = client.get("/api/v1/chat/sessions")
+    assert res.status_code == 401
+
+
+def test_memory_facts_isolated(client: TestClient):
+    res_a = client.get("/api/v1/memory", headers=_headers(client, "a"))
+    res_b = client.get("/api/v1/memory", headers=_headers(client, "b"))
+    assert res_a.status_code == 200
+    assert res_b.status_code == 200
+    facts_a = res_a.json().get("facts") or []
+    facts_b = res_b.json().get("facts") or []
+    assert any("Секрет только для owner" in f.get("content", "") for f in facts_a)
+    assert not any("Секрет только для owner" in f.get("content", "") for f in facts_b)
+    assert res_b.json().get("total_facts", 0) == 0

docker-compose.yml

@@ -1,22 +1,39 @@
-services:
-  backend:
-    build: ./backend
-    ports:
-      - "${BACKEND_PORT:-8080}:${BACKEND_INTERNAL_PORT:-8080}"
-    env_file: .env
-    volumes:
-      - ./data:/app/data
-    extra_hosts:
-      - "host.docker.internal:host-gateway"
-    restart: unless-stopped
-
-  frontend:
-    build:
-      context: ./frontend
-      args:
-        VITE_API_URL: ""
-    ports:
-      - "${FRONTEND_PORT:-3080}:${FRONTEND_INTERNAL_PORT:-80}"
-    depends_on:
-      - backend
-    restart: unless-stopped
+services:
+  qdrant:
+    image: qdrant/qdrant:v1.12.5
+    ports:
+      - "${QDRANT_PORT:-6333}:6333"
+      - "${QDRANT_GRPC_PORT:-6334}:6334"
+    volumes:
+      - qdrant_data:/qdrant/storage
+    restart: unless-stopped
+
+  backend:
+    build: ./backend
+    ports:
+      - "${BACKEND_PORT:-8080}:${BACKEND_INTERNAL_PORT:-8080}"
+    env_file: .env
+    environment:
+      QDRANT_URL: ${QDRANT_URL:-http://qdrant:6333}
+    depends_on:
+      - qdrant
+    volumes:
+      - ./data:/app/data
+    extra_hosts:
+      - "host.docker.internal:host-gateway"
+    restart: unless-stopped
+
+  frontend:
+    build:
+      context: ./frontend
+      args:
+        VITE_API_URL: ""
+        VITE_API_TOKEN: ${VITE_API_TOKEN:-}
+    ports:
+      - "${FRONTEND_PORT:-3080}:${FRONTEND_INTERNAL_PORT:-80}"
+    depends_on:
+      - backend
+    restart: unless-stopped
+
+volumes:
+  qdrant_data:

docker-compose.yml.refactor_bak

@@ -0,0 +1,22 @@
+services:
+  backend:
+    build: ./backend
+    ports:
+      - "${BACKEND_PORT:-8080}:${BACKEND_INTERNAL_PORT:-8080}"
+    env_file: .env
+    volumes:
+      - ./data:/app/data
+    extra_hosts:
+      - "host.docker.internal:host-gateway"
+    restart: unless-stopped
+
+  frontend:
+    build:
+      context: ./frontend
+      args:
+        VITE_API_URL: ""
+    ports:
+      - "${FRONTEND_PORT:-3080}:${FRONTEND_INTERNAL_PORT:-80}"
+    depends_on:
+      - backend
+    restart: unless-stopped

frontend/Dockerfile

@@ -5,6 +5,9 @@ WORKDIR /app
 ARG VITE_API_URL=
 ENV VITE_API_URL=$VITE_API_URL
 
+ARG VITE_API_TOKEN=
+ENV VITE_API_TOKEN=$VITE_API_TOKEN
+
 COPY package.json ./
 RUN npm install
 

frontend/package.json

@@ -1,24 +1,25 @@
-{
-  "name": "home-assistant-frontend",
-  "private": true,
-  "version": "0.1.0",
-  "type": "module",
-  "scripts": {
-    "dev": "vite",
-    "build": "tsc -b && vite build",
-    "preview": "vite preview"
-  },
-  "dependencies": {
-    "react": "^18.3.1",
-    "react-dom": "^18.3.1",
-    "react-markdown": "^9.0.1",
-    "react-router-dom": "^6.28.0"
-  },
-  "devDependencies": {
-    "@types/react": "^18.3.12",
-    "@types/react-dom": "^18.3.1",
-    "@vitejs/plugin-react": "^4.3.3",
-    "typescript": "^5.6.3",
-    "vite": "^5.4.11"
-  }
-}
+{
+  "name": "home-assistant-frontend",
+  "private": true,
+  "version": "0.1.0",
+  "type": "module",
+  "scripts": {
+    "dev": "vite",
+    "build": "node -e \"try{require('fs').unlinkSync('src/pages/Chat.old.tsx')}catch(e){}\" && tsc -b && vite build",
+    "preview": "vite preview"
+  },
+  "dependencies": {
+    "@tanstack/react-virtual": "^3.14.2",
+    "react": "^18.3.1",
+    "react-dom": "^18.3.1",
+    "react-markdown": "^9.0.1",
+    "react-router-dom": "^6.28.0"
+  },
+  "devDependencies": {
+    "@types/react": "^18.3.12",
+    "@types/react-dom": "^18.3.1",
+    "@vitejs/plugin-react": "^4.3.3",
+    "typescript": "^5.6.3",
+    "vite": "^5.4.11"
+  }
+}

frontend/package.json.refactor_bak

@@ -0,0 +1,24 @@
+{
+  "name": "home-assistant-frontend",
+  "private": true,
+  "version": "0.1.0",
+  "type": "module",
+  "scripts": {
+    "dev": "vite",
+    "build": "tsc -b && vite build",
+    "preview": "vite preview"
+  },
+  "dependencies": {
+    "react": "^18.3.1",
+    "react-dom": "^18.3.1",
+    "react-markdown": "^9.0.1",
+    "react-router-dom": "^6.28.0"
+  },
+  "devDependencies": {
+    "@types/react": "^18.3.12",
+    "@types/react-dom": "^18.3.1",
+    "@vitejs/plugin-react": "^4.3.3",
+    "typescript": "^5.6.3",
+    "vite": "^5.4.11"
+  }
+}

frontend/src/App.css

@@ -1,81 +1,134 @@
-.app {
-  height: 100%;
-  min-height: 0;
-  display: flex;
-  flex-direction: column;
-  overflow: hidden;
-}
-
-.app-header {
-  display: flex;
-  align-items: center;
-  justify-content: space-between;
-  padding: 1rem 1.5rem;
-  border-bottom: 1px solid #2a2f3a;
-  background: #151922;
-}
-
-.app-header h1 {
-  margin: 0;
-  font-size: 1.1rem;
-  font-weight: 600;
-}
-
-.app-header nav {
-  display: flex;
-  align-items: center;
-  gap: 0.75rem;
-}
-
-.app-header nav a {
-  padding: 0.45rem 0.9rem;
-  border-radius: 8px;
-  color: #a8b0bd;
-}
-
-.app-header nav a.active {
-  background: #2b3445;
-  color: #fff;
-}
-
-.app-main {
-  flex: 1;
-  min-height: 0;
-  overflow-x: hidden;
-  overflow-y: auto;
-  -webkit-overflow-scrolling: touch;
-  overscroll-behavior: contain;
-}
-
-@media (max-width: 768px) {
-  .app-header {
-    padding: 0.55rem 0.75rem;
-    gap: 0.5rem;
-    flex-shrink: 0;
-  }
-
-  .app-header h1 {
-    display: none;
-  }
-
-  .app-header nav {
-    flex: 1;
-    overflow-x: auto;
-    flex-wrap: nowrap;
-    gap: 0.35rem;
-    padding-bottom: 0.1rem;
-    -webkit-overflow-scrolling: touch;
-    scrollbar-width: none;
-  }
-
-  .app-header nav::-webkit-scrollbar {
-    display: none;
-  }
-
-  .app-header nav a {
-    padding: 0.4rem 0.65rem;
-    font-size: 0.85rem;
-    white-space: nowrap;
-    flex-shrink: 0;
-  }
-}
+.app {
+  height: 100%;
+  min-height: 0;
+  display: flex;
+  flex-direction: column;
+  overflow: hidden;
+}
+
+.app-header {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  padding: 1rem 1.5rem;
+  border-bottom: 1px solid #2a2f3a;
+  background: #151922;
+}
+
+.app-header h1 {
+  margin: 0;
+  font-size: 1.1rem;
+  font-weight: 600;
+}
+
+.app-header nav {
+  display: flex;
+  align-items: center;
+  gap: 0.75rem;
+}
+
+.app-header nav a {
+  padding: 0.45rem 0.9rem;
+  border-radius: 8px;
+  color: #a8b0bd;
+}
+
+  .app-header nav a.active {
+  background: #2b3445;
+  color: #fff;
+}
+
+.app-user {
+  display: inline-flex;
+  align-items: center;
+  gap: 0.5rem;
+  margin-left: 0.25rem;
+  padding-left: 0.5rem;
+  border-left: 1px solid #2a2f3a;
+  color: #8b939f;
+  font-size: 0.8rem;
+  white-space: nowrap;
+}
+
+.app-logout {
+  padding: 0.35rem 0.6rem;
+  border: 1px solid #3a4254;
+  border-radius: 6px;
+  background: transparent;
+  color: #c5ccd6;
+  cursor: pointer;
+  font-size: 0.75rem;
+}
+
+.app-logout:hover {
+  background: #2b3445;
+}
+
+.app-main {
+  position: relative;
+  flex: 1;
+  min-height: 0;
+  overflow-x: hidden;
+  overflow-y: auto;
+  -webkit-overflow-scrolling: touch;
+  overscroll-behavior: contain;
+}
+
+@media (max-width: 768px) {
+  .app-header {
+    padding: 0.55rem 0.75rem;
+    gap: 0.5rem;
+    flex-shrink: 0;
+  }
+
+  .app-header h1 {
+    display: none;
+  }
+
+  .app-header nav {
+    flex: 1;
+    overflow-x: auto;
+    flex-wrap: nowrap;
+    gap: 0.35rem;
+    padding-bottom: 0.1rem;
+    -webkit-overflow-scrolling: touch;
+    scrollbar-width: none;
+  }
+
+  .app-header nav::-webkit-scrollbar {
+    display: none;
+  }
+
+  .app-header nav a {
+    padding: 0.4rem 0.65rem;
+    font-size: 0.85rem;
+    white-space: nowrap;
+    flex-shrink: 0;
+  }
+}
+
+.app-main-chat {
+  overflow: hidden;
+}
+
+.route-panel {
+  position: absolute;
+  inset: 0;
+  display: flex;
+  flex-direction: column;
+  min-height: 0;
+  overflow: hidden;
+  visibility: hidden;
+  pointer-events: none;
+}
+
+.route-panel-active {
+  visibility: visible;
+  pointer-events: auto;
+}
+
+.app-main:not(.app-main-chat) > .route-panel-active {
+  overflow-x: hidden;
+  overflow-y: auto;
+  -webkit-overflow-scrolling: touch;
+}

frontend/src/App.css.refactor_bak

@@ -0,0 +1,81 @@
+.app {
+  height: 100%;
+  min-height: 0;
+  display: flex;
+  flex-direction: column;
+  overflow: hidden;
+}
+
+.app-header {
+  display: flex;
+  align-items: center;
+  justify-content: space-between;
+  padding: 1rem 1.5rem;
+  border-bottom: 1px solid #2a2f3a;
+  background: #151922;
+}
+
+.app-header h1 {
+  margin: 0;
+  font-size: 1.1rem;
+  font-weight: 600;
+}
+
+.app-header nav {
+  display: flex;
+  align-items: center;
+  gap: 0.75rem;
+}
+
+.app-header nav a {
+  padding: 0.45rem 0.9rem;
+  border-radius: 8px;
+  color: #a8b0bd;
+}
+
+.app-header nav a.active {
+  background: #2b3445;
+  color: #fff;
+}
+
+.app-main {
+  flex: 1;
+  min-height: 0;
+  overflow-x: hidden;
+  overflow-y: auto;
+  -webkit-overflow-scrolling: touch;
+  overscroll-behavior: contain;
+}
+
+@media (max-width: 768px) {
+  .app-header {
+    padding: 0.55rem 0.75rem;
+    gap: 0.5rem;
+    flex-shrink: 0;
+  }
+
+  .app-header h1 {
+    display: none;
+  }
+
+  .app-header nav {
+    flex: 1;
+    overflow-x: auto;
+    flex-wrap: nowrap;
+    gap: 0.35rem;
+    padding-bottom: 0.1rem;
+    -webkit-overflow-scrolling: touch;
+    scrollbar-width: none;
+  }
+
+  .app-header nav::-webkit-scrollbar {
+    display: none;
+  }
+
+  .app-header nav a {
+    padding: 0.4rem 0.65rem;
+    font-size: 0.85rem;
+    white-space: nowrap;
+    flex-shrink: 0;
+  }
+}