Develop an automated security audit tool that chains LLM reasoning to perform 'test-driven exploitation'. Focus on generating proof-of-concept scripts for discovered vulnerabilities automatically.
Suggested repo: x-agent
"AI that doesn't just find bugs, it writes the exploits to prove them."
Estimated effort: 80h