TG Agent

A personal Telegram monitoring toolkit — collect messages, search across channels, get keyword alerts. Built as a pet project for my own use.

Русская версия

Features

• All chat types — channels, supergroups, gigagroups, forums, public and private
• Multi-account with automatic flood-wait rotation
• 3 search modes — local DB (FTS5), direct Telegram API, AI/LLM-powered
• AI Agent — interactive chat powered by claude-agent-sdk with automatic deepagents fallback when Claude SDK is not configured; developer override is available in Settings
• All search results are cached in a local SQLite database
• Scheduled collection — incremental message fetching on a timer
• Keyword monitoring — plain text and regex, with Telegram bot notifications
• Built-in anti-spam filters — deduplication, low-uniqueness detection, cross-channel spam, subscriber ratio filters, non-Cyrillic content filter
• Task queue — background job processing with status tracking
• Web dashboard — FastAPI + Bootstrap 5, manage everything from a browser
• Security — session encryption (Fernet + PBKDF2), web panel password, HTTP Basic fallback, HMAC-signed cookies
• Docker-ready

Quick Start

Prerequisites

• Python 3.11+
• Telegram API credentials from my.telegram.org/apps

Installation

pip install tg-agent

Or from source:

pip install .
cp .env.example .env

Edit .env:

TG_API_ID=your_api_id
TG_API_HASH=your_api_hash
WEB_PASS=your_password
SESSION_ENCRYPTION_KEY=    # encrypts account session strings in DB
LLM_API_KEY=               # optional, for AI search
AGENT_MODEL=               # optional, Claude SDK model override
AGENT_FALLBACK_MODEL=      # optional, provider:model for deepagents fallback
AGENT_FALLBACK_API_KEY=    # optional, explicit API key for deepagents fallback provider

Start the server:

python -m src.main serve

Open http://localhost:8080 in your browser and enter the WEB_PASS password.

Docker

cp .env.example .env
# fill in your credentials
docker-compose up -d

Configuration

Environment Variables (.env)

Variable	Required	Description
TG_API_ID	Yes	Telegram API ID
TG_API_HASH	Yes	Telegram API Hash
WEB_PASS	Yes	Web panel password
SESSION_ENCRYPTION_KEY	No*	Key for encrypting Telegram session strings in DB
LLM_API_KEY	No	API key for AI-powered search (deepagents)
ANTHROPIC_API_KEY	No	Anthropic API key for claude-agent-sdk only
CLAUDE_CODE_OAUTH_TOKEN	No	Claude Code auth token for claude-agent-sdk only
AGENT_MODEL	No	Override Claude SDK model for /agent
AGENT_FALLBACK_MODEL	No	provider:model for deepagents fallback in /agent
AGENT_FALLBACK_API_KEY	No	Explicit API key passed to LangChain init_chat_model(...) for fallback

* If not set, sessions are stored in plaintext. If the DB already contains encrypted sessions (enc:v*), startup fails until this key is provided.

config.yaml

Supports ${ENV_VAR} substitution. Empty env vars are dropped (defaults apply).

Section	Description
telegram	API credentials (api_id, api_hash)
web	Host, port, password (default: 0.0.0.0:8080)
scheduler	Collection interval, delays, limits, max flood wait
notifications	admin_chat_id for keyword match alerts
database	SQLite path (default: data/tg_search.db)
llm	LLM provider, model, API key for AI search (deepagents)
agent	Claude model override and deepagents fallback settings for /agent
security	Session encryption settings

Agent backend rules

• /agent uses claude-agent-sdk when ANTHROPIC_API_KEY or CLAUDE_CODE_OAUTH_TOKEN is configured.
• If Claude SDK is not configured, /agent falls back to deepagents when AGENT_FALLBACK_MODEL is set.
• ANTHROPIC_API_KEY and CLAUDE_CODE_OAUTH_TOKEN are never reused by deepagents.
• Developer override for forcing claude-agent-sdk or deepagents lives on the Settings page and applies only when developer mode is enabled.

CLI

# Web server
python -m src.main [--config CONFIG] serve [--web-pass PASS]
python -m src.main [--config CONFIG] stop
python -m src.main [--config CONFIG] restart [--web-pass PASS]

# One-shot collection
python -m src.main [--config CONFIG] collect [--channel-id ID]

# Search
python -m src.main [--config CONFIG] search "query" [--limit N] [--mode MODE]

# Channel management
python -m src.main channel list|add|delete|toggle|collect|stats|refresh-types|import

# Content filters
python -m src.main filter analyze|apply|reset|precheck

# Keywords
python -m src.main keyword list|add|delete|toggle

# Accounts
python -m src.main account list|toggle|delete

# Scheduler
python -m src.main scheduler start|trigger|search

# Notification bot
python -m src.main notification setup|status|delete

# Diagnostics and benchmarks
python -m src.main test all|read|write|telegram|benchmark

telethon-cli

telethon-cli is installed with the project and reuses the same TG_API_ID and TG_API_HASH values from .env.

Optional CLI-only overrides:

• TG_SESSION sets a custom Telethon session path or name.
• TG_PASSWORD supplies the Telegram 2FA password for non-interactive runs.

Legacy TELETHON_* environment variable names are still accepted by telethon-cli for compatibility, but this project standardizes on TG_*.

telethon-cli login
telethon-cli users get-me --output json

Web Interface

Page	Path	Description
Web login	/login	Sign in to the web panel with WEB_PASS
Dashboard	/	Stats, scheduler status, connected accounts
Telegram auth	/auth/login	Add Telegram accounts (phone + code + 2FA)
Accounts	/accounts	Manage connected accounts
Channels	/channels	Add/remove channels, keywords, import
Search	/search	Search messages (local / Telegram / AI)
Analytics	/analytics	Top posts leaderboard, engagement by content type, hourly patterns
Filters	/filter	Anti-spam filter report and controls
Scheduler	/scheduler	Start/stop/trigger collection and keyword search
Agent	/agent	AI chat assistant with access to collected messages

Roadmap

• LLM-powered content factory
• LLM-powered intelligent search
• LLM-based chat spam moderation
• Direct message handling
• Telegram action automation (broadcasts, etc.)

Development

# Install dev dependencies
pip install -e ".[dev]"

# Run parallel-safe tests (all available CPUs minus one worker)
pytest tests/ -v -m "not aiosqlite_serial" -n auto

# Run aiosqlite-backed tests serially
pytest tests/ -v -m aiosqlite_serial

# Run a single test
pytest tests/test_web.py::test_health_endpoint -v

# Benchmark serial vs safe mixed-mode suite execution
python -m src.main test benchmark

# Lint
ruff check src/ tests/ conftest.py