TLDR Tech / DevOps
security-audit
A new coding-agent skill turns an AI agent into a security auditor that orchestrates reconnaissance, vulnerability hunting, and validation. The tool runs a structured audit in six phases and classifies findings into three verdicts: confirmed, needs_validation, or rejected. Audits against the same repository are additive, using prior results to target gaps or revalidate changed source code. The skill activates automatically when an agent receives a request for a security audit or pen-test.
Read the full story