feat: add conversational Telegram AI chat

README.md

@@ -164,6 +164,11 @@ TELEGRAM_BOT_TOKEN=
 TELEGRAM_CHAT_ID=
 TELEGRAM_POLL_INTERVAL=5000
 TELEGRAM_CHANNELS=
+TELEGRAM_AI_CHAT_ENABLED=true
+TELEGRAM_AI_HISTORY_MESSAGES=8
+TELEGRAM_AI_MAX_INPUT_CHARS=2000
+TELEGRAM_AI_MAX_TOKENS=2048
+TELEGRAM_AI_TIMEOUT_MS=300000
 DISCORD_BOT_TOKEN=
 DISCORD_CHANNEL_ID=
 DISCORD_GUILD_ID=
@@ -360,13 +365,17 @@ Intelligence Terminal doubles as an interactive Telegram bot. Beyond sending ale
 | `/status` | System health, last sweep time, source status, LLM status |
 | `/sweep` | Trigger a manual sweep cycle |
 | `/brief` | Compact text summary of the latest intelligence (direction, key metrics, top OSINT) |
+| `/ask <question>` | Ask the configured LLM about the latest intelligence and conversation context |
+| `/reset` | Clear the in-memory AI conversation history |
 | `/portfolio` | Portfolio status (if Alpaca connected) |
 | `/alerts` | Recent alert history with tiers |
 | `/mute` / `/mute 2h` | Silence alerts for 1h (or custom duration) |
 | `/unmute` | Resume alerts |
 | `/help` | Show all available commands |
 
-This requires `TELEGRAM_BOT_TOKEN` and `TELEGRAM_CHAT_ID` in `.env`. The bot polls for messages every 5 seconds (configurable via `TELEGRAM_POLL_INTERVAL`).
+Normal text messages in the configured private chat are treated as AI questions, so commands are optional. Answers include a compact snapshot of the latest sweep, recent ideas, evidence links, degraded sources, and a bounded conversation history. Snapshot fields are treated as untrusted evidence rather than instructions. Conversation history remains in memory only and is cleared on restart or with `/reset`.
+
+This requires `TELEGRAM_BOT_TOKEN`, `TELEGRAM_CHAT_ID`, and a configured LLM in `.env`. The bot ignores every other chat ID and polls every 5 seconds by default. For group chats, BotFather privacy settings may prevent the bot from receiving normal text; private chat is recommended. Local reasoning models can take several minutes, so the bot refreshes Telegram's typing indicator while waiting.
 
 ### Discord Bot (Two-Way)
 
@@ -660,6 +669,11 @@ All settings are in `.env` with sensible defaults:
 | `TELEGRAM_CHAT_ID` | — | Your Telegram chat ID |
 | `TELEGRAM_CHANNELS` | — | Extra channel IDs to monitor (comma-separated) |
 | `TELEGRAM_POLL_INTERVAL` | `5000` | Bot command polling interval (ms) |
+| `TELEGRAM_AI_CHAT_ENABLED` | `true` | Reply to normal Telegram text with the configured LLM |
+| `TELEGRAM_AI_HISTORY_MESSAGES` | `8` | Maximum user/assistant messages retained in memory |
+| `TELEGRAM_AI_MAX_INPUT_CHARS` | `2000` | Maximum characters accepted from one Telegram message |
+| `TELEGRAM_AI_MAX_TOKENS` | `2048` | Maximum tokens for one Telegram AI answer |
+| `TELEGRAM_AI_TIMEOUT_MS` | `300000` | Telegram AI request timeout for local models |
 | `DISCORD_BOT_TOKEN` | disabled | For Discord alerts + slash commands |
 | `DISCORD_CHANNEL_ID` | — | Discord channel for alerts |
 | `DISCORD_GUILD_ID` | — | Server ID (instant slash command registration) |
@@ -738,7 +752,7 @@ OpenSky can also return `HTTP 429` when its public hotspots are queried too aggr
 
 ### Telegram bot not responding to commands
 
-Make sure both `TELEGRAM_BOT_TOKEN` and `TELEGRAM_CHAT_ID` are set in `.env`. The bot only responds to messages from the configured chat ID (security measure). You should see Telegram alert and bot polling startup lines in the server logs. If not, double-check your token with `curl https://api.telegram.org/bot<YOUR_TOKEN>/getMe`.
+Make sure `TELEGRAM_BOT_TOKEN`, `TELEGRAM_CHAT_ID`, and the LLM settings are present in `.env`. The bot only responds to the configured chat ID. Check `/status` for LLM and AI-chat state, then try `/ask What changed?`. You should see Telegram polling startup lines in the server logs. If command registration fails, verify the bot token without posting it publicly. In group chats, use `/ask` or adjust BotFather privacy settings because normal text may not be delivered to the bot.
 
 ### Discord bot not responding to slash commands