THE MOST POWERFUL OFFENSIVE AI AGENT FOR PENTESTING , OSINT , AND ETHICAL HACKING
-
Updated
Oct 28, 2026 - Python
THE MOST POWERFUL OFFENSIVE AI AGENT FOR PENTESTING , OSINT , AND ETHICAL HACKING
Toolkit for AI whitehats, internal red teams, llm bug bounty hunters and mlops - Adversarial testing for AI, LLMs, and Agents.
Burp MCP Security Analysis Toolkit
Mergen is an MCP server that gives your AI a real red team brain. It doesn't just run tools, it picks the right ones, chains them together, and actually makes sense of the output. Built by pentesters, for pentesters who are tired of babysitting scripts.
Claude Code Harness
Protect AI — independent third-party profile of a public API surface, by API Evangelist. Protect AI is an AI/ML security company building defensive tooling and open-source security research for the model supply chain, large language model (LLM) applications, and runtime AI workloads.
To associate your repository with the ai-bug-bounty topic, visit your repo's landing page and select "manage topics."