Back to the chart

cloudflare/security-audit-skill

Multi-phase security audit skill for coding agents

Monthly
#2
Stars
75.5k

Open on GitHub

What it is

A coding-agent skill that orchestrates isolated agents through a six-phase security audit covering reconnaissance, coverage-led hunting, candidate validation, structured output, independent verification, and reporting. Verified records are written to findings.json and derived into target-neutral Markdown reports.

Who it is for

Developers and security practitioners using coding agents with tool use and parallel sub-agents to audit codebases for vulnerabilities.

Why it is trending

It gained 43,924 stars today, reaching 74,364 total stars. Seeded from Cloudflare's vulnerability discovery harness, it separates finding agents from verifying agents and supports additive runs across the same repository.

Main features

  • Six-phase audit workflow from reconnaissance to target-neutral reporting
  • Adversarial validation using separate agents to disprove candidate findings
  • Zero-dependency Node.js validators for findings.json and coverage-ledger.json
  • Domain-specific hunting guides for web, cloud, AI/LLM, binary, and supply-chain targets

Summarized on November 17, 2026 from the README and repository data. It can be wrong; check the source before relying on it. Report a wrong summary