English | 한국어

A lighter, easier alternative to OpenClaw/Hermes.

Runs autonomously inside Docker and chats with you through Telegram.

What it does

• Daily chat — Send messages and get replies right in Telegram. Supports text, images, and files.
• Connect your inference provider — Works with OpenAI, OpenRouter, or any custom API endpoint.
• Skills, MCP — Add capabilities like web browsing, scheduled tasks, and more.
• Scheduled tasks — Set up recurring jobs that run automatically and report back.
• Runs anywhere — Lightweight single Docker container, works on any device.

Quick start

1. Create a Telegram bot

1. Open Telegram and search for @BotFather.
2. Send /newbot and follow the prompts.
3. Copy the bot token you receive.

2. Install (Linux / macOS)

Paste this one line into your terminal and press Enter. The installer will ask you to paste the full token from BotFather, then (optionally) whether to connect a host folder — default no; if yes, it warns about risk and asks for an absolute path.

When the installer asks you to install Docker, press y to install it. tabyAgent requires Docker to run. Automatic installation is supported on macOS and Linux only.

curl -fsSL https://raw.githubusercontent.com/gpdir16/tabyAgent/main/scripts/install.sh | bash

When it finishes, open your bot in Telegram and send /start. To upgrade later, run the same command again — your settings and memory are kept.

3. Configure in Telegram

1. Open your bot in Telegram and send /start. The first person to send a message is automatically approved.
2. The setup wizard will guide you through language, LLM provider, API key, and model.
3. Once done, you can start chatting.

Send /config anytime to change your settings.

Run from source (Docker Compose)

git clone https://github.com/gpdir16/tabyAgent.git
cd tabyAgent
cp .env.example .env   # set TELEGRAM_BOT_TOKEN
docker compose up -d

Optional host folder (default: none):

echo 'HOST_WORKSPACE=/absolute/path/to/your/project' >> .env
docker compose -f docker-compose.yml -f docker-compose.workspace.yml up -d

Mounts at /workspace in the container. Default work stays under /app/user; use /workspace only when a task needs files on the user's PC.

Rebuild the image after pulling code changes:

docker compose up -d --build

Bot commands

Requirements

• Docker
• Telegram bot token (from BotFather)
• Inference provider API key

License

AGPL-3.0