This repository serves as a curated, evolving list of open-source AI hacking agents, targeting security researchers and developers. It aims to centralize information on these agents, facilitating their discovery, comparison, and integration into security workflows, thereby accelerating the development and adoption of AI-powered offensive security tools.

How It Works

The project compiles links to various AI hacking agent repositories, often including direct links to associated "DeepWiki" and "CodeWiki" documentation. A key feature is the integration with the DeepWiki MCP, a public server that exposes structured tools (like ask_question, read_wiki_structure, read_wiki_contents) enabling AI agents to query and understand the features of other open-source repositories. This facilitates agent-to-agent knowledge sharing and development.

Quick Start & Requirements

• This is a curated list, not a software package to install directly.
• Integration with the DeepWiki MCP (https://mcp.deepwiki.com/mcp) is facilitated by adding the MCP endpoint to compatible AI IDEs/Agents (e.g., Cursor, Claude Code, Windsurf, Continue.dev).
• Usage Disclaimer: "THESE REPOS ARE FOR EDUCATIONAL AND AUTHORIZED SECURITY TESTING PURPOSES ONLY."

Highlighted Details

• Maintains a list of approximately 64 open-source AI hacking agents, with plans to expand and refine categorization.
• Features integration with the DeepWiki MCP for querying open-source repo details.
• Includes links to DARPA AIxCC finalist repositories and Tencent Challenge participants.
• Work in progress: Benchmark results, detailed feature analysis, and sorting are pending.

Maintenance & Community

The repository is explicitly marked as a "WORK IN PROGRESS PLEASE CONTRIBUTE!!". Contributions are encouraged via GitHub issues for missing agents. A Discord community is available at https://discord.gg/9AJnkNe6RE.

Licensing & Compatibility

No specific license is listed for the curated list itself. However, a strong disclaimer mandates that the listed repositories are for "EDUCATIONAL AND AUTHORIZED SECURITY TESTING PURPOSES ONLY." Compatibility is focused on integration with AI development tools that support the DeepWiki MCP.

Limitations & Caveats

The list is incomplete and actively under development, with many entries pending further details like benchmark scores and refined categorization. The strict usage disclaimer limits the application of the listed tools to educational and authorized security testing contexts.