XAgent

Active
GitHub Python Apache-2.0

Description

An autonomous LLM agent framework for complex task solving with automatic task decomposition, tool usage, and multi-step reasoning from the OpenBMB team

Key Features

  • Autonomous task decomposition and planning: Planner automatically breaks complex tasks into subtasks with execution milestones
  • Secure sandbox execution: all operations run inside Docker containers for environment isolation and safety
  • Extensible tool set: ToolServer provides file editing, Python Notebook, web browser, shell, and Rapid API tools
  • Human-AI collaboration: agents proactively seek human assistance when encountering challenges, or follow human guidance step-by-step
  • Dynamic Dispatcher scheduling: instantiates and dispatches tasks to different agents dynamically, extensible with new agents
  • Web GUI interface: provides both a visual web interface and command-line interaction

Use Cases

💡 Automated data analysis tasks: executing data cleaning, visualization, and report generation autonomously
💡 Software development assistance: completing code writing and debugging through Python Notebook and shell tools
💡 Web information retrieval and integration: automatically searching, browsing, and summarizing web content via the web browser tool
💡 API integration testing: calling external APIs through Rapid API tools and validating response results

Quick Start

1. Install Docker and Docker Compose
2. Run: docker compose up to start ToolServer
3. Run: pip install -r requirements.txt to install Python dependencies
4. Configure OpenAI API Key in assets/config.yml
5. Run: python run.py --task "your task description"
6. Visit http://localhost:5173 to use the Web GUI

Related Projects