# auto-agent

**Repository Path**: mkwchecking/auto-agent

## Basic Information

- **Project Name**: auto-agent
- **Description**: 个人用多 agent 协作库
- **Primary Language**: Python
- **License**: MIT
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2026-02-05
- **Last Updated**: 2026-03-30

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# Auto-Agent

A multi-agent system based on MAF (Microsoft Agent Framework).

## Features

- **Multi-channel Support**: Telegram, Discord, Slack, Feishu, DingTalk, WebSocket, Terminal
- **Multi-agent Collaboration**: Handoff workflow for agent coordination
- **Memory System**: Short-term and long-term memory with context engineering
- **Skill System**: SKILL.md files for skill definition
- **Tool System**: @tool decorator for tool registration
- **Scheduled Tasks**: Cron and Heartbeat for task scheduling
- **Approval System**: Safe handling of sensitive operations
- **LLM Support**: Azure OpenAI, OpenAI, Anthropic, Ollama

## Requirements

- Python >= 3.11
- Operating System: Linux, macOS, or Windows

## Installation

### Option 1: Using Installation Script (Recommended)

**Linux/macOS:**
```bash
./install.sh
```

**Windows:**
```cmd
install.bat
```

The script will:
- Check for existing Python installations
- Use system Python (>= 3.11) if available
- Install Miniconda only if no suitable Python is found
- Create a virtual environment and install dependencies

### Option 2: Manual Installation

```bash
# Create virtual environment
python -m venv .venv
source .venv/bin/activate  # Linux/macOS
# or
.venv\Scripts\activate  # Windows

# Install package
pip install -e .
```

## Quick Start

### Using run.sh (Linux/macOS)

The `run.sh` script provides a convenient way to run the application:

```bash
# CLI mode (interactive chat)
./run.sh cli

# CLI with additional arguments
./run.sh cli chat
./run.sh cli config
./run.sh cli status

# API server mode
./run.sh api

# API server with custom host/port
AUTO_AGENT_HOST=127.0.0.1 AUTO_AGENT_PORT=3000 ./run.sh api
```

### CLI Mode

```bash
# Interactive chat
python -m auto_agent.app chat

# Configuration management
python -m auto_agent.app config

# Show system status
python -m auto_agent.app status
```

### Server Mode

```bash
# Start backend server
python -m auto_agent.app server --api-port 8000 --ws-port 8765
```

### Programmatic Usage

```python
from auto_agent import Application

app = Application()
await app.start()
```

## Project Structure

```
auto-agent/
├── src/auto_agent/
│   ├── app/                    # Application layer
│   │   ├── api/                # REST API routes
│   │   ├── cli.py              # CLI commands
│   │   └── server.py           # Backend server
│   ├── core/                   # Core components
│   │   ├── agent/              # Agent definitions
│   │   ├── llm/                # LLM providers
│   │   ├── memory/             # Memory system
│   │   ├── runner/             # Execution runner
│   │   ├── skill/              # Skill system
│   │   └── tool/               # Tool registry
│   └── ext/                    # Extensions
│       ├── channel/            # Channel plugins
│       ├── config/             # Configuration
│       └── cron/               # Scheduling
├── tests/                      # Test suite
├── install.sh                  # Unix installation script
├── install.bat                 # Windows installation script
└── pyproject.toml              # Project configuration
```

## CLI Commands

| Command | Description |
|---------|-------------|
| `chat` | Start interactive chat session |
| `config` | Interactive configuration |
| `status` | Show system status |
| `channel start` | Start all channels |
| `channel stop` | Stop all channels |
| `channel list` | List all channels |
| `llm list` | List LLM providers |
| `llm activate <name>` | Activate LLM provider |

## Configuration

Configuration is stored in `.auto-agent/config.db` (SQLite).

Key configuration items:
- `default_workflow`: Default workflow name
- `websocket_host`: WebSocket server host
- `websocket_port`: WebSocket server port

## Development

```bash
# Run tests
python -m pytest tests/ -v

# Run linter
ruff check src/

# Type check
mypy src/
```

## Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (`git checkout -b feature/amazing-feature`)
3. Commit your changes (`git commit -m 'Add some amazing feature'`)
4. Push to the branch (`git push origin feature/amazing-feature`)
5. Open a Pull Request

## License

This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.