dinosn/raptor-loop-hunt

This skill implements a recursive auto-research methodology for security researchers conducting exhaustive audits of codebases. By employing a generate-then-judge discipline, it moves beyond single-pass scans to identify exploitable bugs through iterative analysis of repositories, services, and modules.

It is designed for scenarios requiring maximum coverage where the goal is to verify vulnerabilities rather than guess them, systematically auditing source code file by file and function by function.