Skip to content

[GitHub Trending] Unclecheng-li/VulnClaw

6.9 relevance
Score Breakdown
technical depth
7
novelty
8
actionability
6
community
6
strategic
6
personal
8

Scored daily by a customisable AI persona to surface the most relevant engineering leadership news.

AI Agent + MCP for penetration testing, directly matches agent orchestration and security interests.

AI/ML github.com
基于 AI Agent + MCP 工具链 + 渗透 Skill 编排, 配合大语言模型, 自然语言输入 → 自动完成「信息收集 → 漏洞发现 → 漏洞利用 → 报告生成」全流程。 - Unclecheng-li/VulnClaw
Summary

VulnClaw is an open-source, AI-driven penetration testing CLI tool that uses LLM agents and MCP toolchains to automate the full pentest workflow from information gathering to report generation. Its core innovation is a goal-driven solver engine that models penetration as a state-space search over Facts and Intents, structurally preventing loops, and includes an evidence-level anti-hallucination gate that rejects any claimed flag not found verbatim in real tool output. The tool supports 13 LLM providers, 21 penetration skills, a plugin system, and a Web UI, with Docker deployment available.

Author

Unclecheng-li