feat: add chat group daily summarizer

- Telegram collector (Telethon MTProto) with checkpoint-based message fetching
- Claude API summarization via proxy (haiku model)
- Dashboard page with date picker, markdown summary, raw message drill-down
- Separate container lifecycle from dashboard for stable Telethon sessions
- Shared SQLite DB mounted read-only into dashboard

chat-summarizer/Dockerfile

@@ -0,0 +1,6 @@
+FROM python:3.12-slim
+WORKDIR /app
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+COPY . .
+CMD ["python", "-u", "main.py"]

chat-summarizer/collector.py

@@ -0,0 +1,80 @@
+import os
+import logging
+import aiosqlite
+from telethon import TelegramClient
+from db import DB_PATH, init_db
+
+log = logging.getLogger(__name__)
+
+API_ID = int(os.environ.get("TG_API_ID", 0))
+API_HASH = os.environ.get("TG_API_HASH", "")
+SESSION_NAME = os.environ.get("TG_SESSION_NAME", "chat_summarizer")
+GROUP_IDS = [int(g) for g in os.environ.get("TG_GROUP_IDS", "").split(",") if g.strip()]
+
+def get_client():
+    return TelegramClient(f"/app/data/{SESSION_NAME}", API_ID, API_HASH)
+
+async def auth():
+    """Interactive one-time auth. Run: docker exec -it chat-summarizer python3 -c 'import asyncio; from collector import auth; asyncio.run(auth())'"""
+    client = get_client()
+    await client.start()
+    me = await client.get_me()
+    print(f"Logged in as {me.first_name} (id={me.id})")
+    await client.disconnect()
+
+async def collect_messages():
+    if not GROUP_IDS:
+        log.warning("No TG_GROUP_IDS configured, skipping collection")
+        return
+
+    client = get_client()
+    await client.connect()
+    if not await client.is_user_authorized():
+        log.error("Not authorized — run interactive auth first")
+        return
+
+    try:
+        async with aiosqlite.connect(DB_PATH) as db:
+            for gid in GROUP_IDS:
+                try:
+                    # Get checkpoint
+                    cursor = await db.execute("SELECT last_msg_id FROM checkpoints WHERE group_id=?", (gid,))
+                    row = await cursor.fetchone()
+                    min_id = row[0] if row else 0
+
+                    entity = await client.get_entity(gid)
+                    group_name = getattr(entity, 'title', str(gid))
+                    count = 0
+
+                    async for msg in client.iter_messages(entity, min_id=min_id, limit=1000):
+                        if not msg.text:
+                            continue
+                        sender = ""
+                        if msg.sender:
+                            sender = getattr(msg.sender, 'first_name', '') or getattr(msg.sender, 'title', '') or ''
+                            last = getattr(msg.sender, 'last_name', '')
+                            if last:
+                                sender += f" {last}"
+
+                        await db.execute(
+                            "INSERT INTO messages (group_id, group_name, sender_name, text, msg_id, timestamp) VALUES (?,?,?,?,?,?)",
+                            (gid, group_name, sender, msg.text, msg.id, msg.date.isoformat())
+                        )
+                        count += 1
+
+                    if count > 0:
+                        # Update checkpoint to max msg_id
+                        cursor = await db.execute("SELECT MAX(msg_id) FROM messages WHERE group_id=?", (gid,))
+                        max_row = await cursor.fetchone()
+                        if max_row and max_row[0]:
+                            await db.execute(
+                                "INSERT OR REPLACE INTO checkpoints (group_id, last_msg_id) VALUES (?,?)",
+                                (gid, max_row[0])
+                            )
+                        log.info("Collected %d messages from %s", count, group_name)
+
+                    await db.commit()
+                except Exception as e:
+                    log.error("Error collecting from group %s: %s", gid, e)
+    finally:
+        await client.disconnect()

chat-summarizer/db.py

@@ -0,0 +1,33 @@
+import aiosqlite
+import os
+
+DB_PATH = os.environ.get("DB_PATH", "/app/data/chat_summary.db")
+
+async def init_db():
+    async with aiosqlite.connect(DB_PATH) as db:
+        await db.execute("""
+            CREATE TABLE IF NOT EXISTS messages (
+                id INTEGER PRIMARY KEY AUTOINCREMENT,
+                group_id INTEGER NOT NULL,
+                group_name TEXT,
+                sender_name TEXT,
+                text TEXT,
+                msg_id INTEGER,
+                timestamp DATETIME NOT NULL
+            )
+        """)
+        await db.execute("""
+            CREATE TABLE IF NOT EXISTS summaries (
+                id INTEGER PRIMARY KEY AUTOINCREMENT,
+                date TEXT UNIQUE NOT NULL,
+                content TEXT NOT NULL,
+                created_at DATETIME DEFAULT CURRENT_TIMESTAMP
+            )
+        """)
+        await db.execute("""
+            CREATE TABLE IF NOT EXISTS checkpoints (
+                group_id INTEGER PRIMARY KEY,
+                last_msg_id INTEGER NOT NULL
+            )
+        """)
+        await db.commit()

chat-summarizer/docker-compose.yml

@@ -0,0 +1,20 @@
+services:
+  chat-summarizer:
+    build: .
+    container_name: chat-summarizer
+    restart: unless-stopped
+    volumes:
+      - /volume1/docker/chat-summarizer/data:/app/data
+    environment:
+      - TG_API_ID=${TG_API_ID}
+      - TG_API_HASH=${TG_API_HASH}
+      - TG_SESSION_NAME=${TG_SESSION_NAME:-chat_summarizer}
+      - TG_GROUP_IDS=${TG_GROUP_IDS}
+      - ANTHROPIC_API_KEY=${ANTHROPIC_API_KEY}
+      - ANTHROPIC_BASE_URL=${ANTHROPIC_BASE_URL:-https://www.bytecatcode.org}
+      - SUMMARY_HOUR=${SUMMARY_HOUR:-23}
+    logging:
+      driver: json-file
+      options:
+        max-size: "10m"
+        max-file: "3"

chat-summarizer/main.py

@@ -0,0 +1,41 @@
+import asyncio
+import logging
+import os
+from datetime import datetime, timedelta, timezone
+
+from db import init_db
+from collector import collect_messages
+from summarizer import summarize
+
+logging.basicConfig(level=logging.INFO, format="%(asctime)s %(name)s %(levelname)s %(message)s")
+log = logging.getLogger(__name__)
+
+CST = timezone(timedelta(hours=8))
+SUMMARY_HOUR = int(os.environ.get("SUMMARY_HOUR", 23))
+COLLECT_INTERVAL = 300  # 5 minutes
+
+async def main():
+    await init_db()
+    log.info("Chat summarizer started (summary hour=%d CST)", SUMMARY_HOUR)
+
+    last_summary_date = None
+
+    while True:
+        try:
+            await collect_messages()
+        except Exception as e:
+            log.error("Collection error: %s", e)
+
+        now = datetime.now(CST)
+        today = now.strftime("%Y-%m-%d")
+        if now.hour >= SUMMARY_HOUR and last_summary_date != today:
+            try:
+                await summarize(today)
+                last_summary_date = today
+            except Exception as e:
+                log.error("Summary error: %s", e)
+
+        await asyncio.sleep(COLLECT_INTERVAL)
+
+if __name__ == "__main__":
+    asyncio.run(main())

chat-summarizer/requirements.txt

@@ -0,0 +1,4 @@
+telethon
+httpx
+aiosqlite
+cryptg

chat-summarizer/summarizer.py

@@ -0,0 +1,77 @@
+import os
+import logging
+from datetime import datetime, timedelta, timezone
+import aiosqlite
+import httpx
+from db import DB_PATH
+
+log = logging.getLogger(__name__)
+
+API_KEY = os.environ.get("ANTHROPIC_API_KEY", "")
+BASE_URL = os.environ.get("ANTHROPIC_BASE_URL", "https://www.bytecatcode.org")
+MODEL = "claude-haiku-4-5-20251001"
+CST = timezone(timedelta(hours=8))
+
+PROMPT = """You are a chat group digest assistant. Summarize the following chat messages into a concise daily digest.
+
+Rules:
+- Group by topic/theme, not by chat group
+- Highlight key decisions, shared links, and action items
+- Skip greetings, small talk, and low-value chatter
+- Use markdown formatting
+- Write in the same language as the messages (likely Chinese)
+- Be concise
+
+Messages:
+{messages}"""
+
+async def summarize(date_str: str = None):
+    now = datetime.now(CST)
+    if not date_str:
+        date_str = now.strftime("%Y-%m-%d")
+
+    start = datetime.strptime(date_str, "%Y-%m-%d").replace(tzinfo=CST)
+    end = start + timedelta(days=1)
+
+    async with aiosqlite.connect(DB_PATH) as db:
+        cursor = await db.execute(
+            "SELECT group_name, sender_name, text, timestamp FROM messages WHERE timestamp >= ? AND timestamp < ? ORDER BY timestamp",
+            (start.isoformat(), end.isoformat())
+        )
+        rows = await cursor.fetchall()
+
+    if not rows:
+        log.info("No messages for %s, skipping summary", date_str)
+        return None
+
+    # Format messages for the prompt
+    formatted = []
+    for group_name, sender, text, ts in rows:
+        formatted.append(f"[{group_name}] {sender}: {text}")
+    messages_text = "\n".join(formatted)
+
+    # Call Claude API
+    async with httpx.AsyncClient(timeout=120) as client:
+        resp = await client.post(
+            f"{BASE_URL}/v1/messages",
+            headers={"x-api-key": API_KEY, "anthropic-version": "2023-06-01", "content-type": "application/json"},
+            json={
+                "model": MODEL,
+                "max_tokens": 4096,
+                "messages": [{"role": "user", "content": PROMPT.format(messages=messages_text)}]
+            }
+        )
+        resp.raise_for_status()
+        data = resp.json()
+        summary = data["content"][0]["text"]
+
+    # Store summary
+    async with aiosqlite.connect(DB_PATH) as db:
+        await db.execute(
+            "INSERT OR REPLACE INTO summaries (date, content, created_at) VALUES (?,?,?)",
+            (date_str, summary, now.isoformat())
+        )
+        await db.commit()
+
+    log.info("Generated summary for %s (%d messages)", date_str, len(rows))
+    return summary