LibreFang
LibreFang is an open-source Agent Operating System written in Rust. It provides full lifecycle management for AI agents — from creation and scheduling to communication and deployment — with 49 LLM drivers, 44 channel adapters connecting platforms like Telegram, Discord, Slack, and WeChat, plus MCP and A2A protocols for inter-agent collaboration.
Manage agents via CLI or the web dashboard, configure workflows and cron jobs, and let agents autonomously browse the web, call APIs, and run commands through Hands. All operations are protected by 16 layers of security with RBAC and audit logging.
Website: https://librefang.ai/
Deploy: https://deploy.librefang.ai/
GitHub: https://github.com/librefang/librefang
Issues: https://github.com/librefang/librefang/issues
Pull Requests: https://github.com/librefang/librefang/pulls
Discord: https://discord.gg/DzTYqAZZmc
Discussions: https://github.com/librefang/librefang/discussions
Quick Links
- Getting Started — Install, configure, and run your first Agent
- GitHub — Source code
Features
15 Hands
Browser, Researcher, Clip, Lead, Collector, Predictor, Creator, DevOps, Trader, and more
16-Layer Security
WASM Sandbox, Merkle Hash Chain, Taint Tracking, SSRF Protection
44 Channel Adapters
Telegram, Discord, Slack, Email, Webhooks, and more
49 LLM Drivers
OpenAI, Anthropic, Gemini, Groq, DeepSeek, Ollama, and more
MCP & A2A Protocols
Model Context Protocol & Agent-to-Agent communication
Open Source
100% open source, Rust-based, self-hostable