LibreAI / Agentes de IA / Security Audit Skill
// agents · cloudflare/security-audit-skill

Security Audit Skill

Security Audit Skill es un proyecto de agentes de IA de código abierto de cloudflare (cloudflare/security-audit-skill en GitHub). Tiene 67.594 estrellas en GitHub y 4.423 bifurcaciones, se publica bajo la licencia MIT y se actualizó por última vez en GitHub el 14 de septiembre de 2026.

Cloudflare's security-audit skill — turns your coding agent into a multi-phase auditor with independently verified findings, MIT

★ 67,6k estrellas⑂ 4,4k bifurcacionesMITactualizado sep 2026
▦ visual: el README de este proyecto no incluye ninguna captura incrustable,
así que aquí va un marcador. La guía de abajo está completa de todos modos.
// datos del proyecto

Datos del proyecto

Estrellas en GitHub
67.594
Bifurcaciones
4.423
Licencia
MIT
Última actualización en GitHub
14 de septiembre de 2026
Categoría
Agentes de IA
Tipo de ficha
Guía de instalación completa
// cómo funciona

Cómo funciona

  1. Install the skill

    Use the Skills CLI to add the security-audit skill from the repo.

  2. Point the agent at the code

    Start your coding agent in (or pointed at) the codebase you want audited.

  3. Ask for an audit

    Say "security audit this codebase" and let the six phases run; read REPORT.md and findings.json when it finishes.

// características

Por qué gusta tanto

Six-phase audit

Reconnaissance, coverage-led hunting, candidate validation, structured output, independent verification, target-neutral reporting.

Findings that survive challenge

Every candidate goes to a fresh verifier that tries to disprove it; only confirmed findings carry a severity.

Machine-readable output

findings.json with confirmed / needs_validation / rejected verdicts, validated against report-schema.json.

Coverage ledger

A ledger records what was checked, so gaps are visible and later runs target exactly what is missing.

Attack-class guides

Hunting prompts for web, memory-safety and binary, LLM and prompt-injection, cloud, supply-chain, protocols, client-side and more.

// ejecútalo

Ejecútalo tú mismo

  1. Step 1 — Install the skill

    Add it with the Skills CLI.

    npx skills add https://github.com/cloudflare/security-audit-skill --skill security-audit
  2. Step 2 — Install globally

    Add --global for a user-level install instead of a per-project one.

    npx skills add https://github.com/cloudflare/security-audit-skill --skill security-audit --global
  3. Step 3 — Run the audit

    In your coding agent, pointed at the target codebase, ask for the audit in plain language.

    security audit this codebase
// casos de uso

Qué puedes crear

  • Audit a codebase before a release
  • Hunt for vulnerabilities with coverage tracking
  • Verify security findings independently before acting on them
  • Learn a structured audit methodology from the phase guides

Two more from the same shelf: Seedance2-Storyboard-Generator — Open skill that turns novels and stories into, and agentic_coding_flywheel_setup — Turn a fresh Ubuntu VPS into a multi-agent dev. Browse every AI Agents tool in the catalogue to compare the category. Before running a model locally, check how much VRAM you need for local LLMs — model size has to match your hardware.

// preguntas frecuentes

Preguntas frecuentes

What does the skill do?

It turns a coding agent into a security auditor that runs a six-phase audit and writes verified, machine-readable findings.

Is it really from Cloudflare?

Yes — the README says it seeded Cloudflare’s vulnerability discovery harness, described in their "Build your own vulnerability harness" post.

What do I need to run it?

A coding agent and the Skills CLI (npx skills). The validators are zero-dependency scripts.

What do the verdicts mean?

confirmed has a complete source trace and bounded observed result; needs_validation has one exact unresolved fact and no severity; rejected was disproved.

Is it free?

Yes — MIT licensed.

// sigue explorando

Herramientas similares

Agentes de IA

Seedance2-Storyboard-Generator

Open skill that turns novels and stories into multi-episode video storyboards.

★ 2,5k estrellasCustom
Ver la guía →
Agentes de IA

agentic_coding_flywheel_setup

Turn a fresh Ubuntu VPS into a multi-agent dev environment.

★ 1,7k estrellasNOASSERTION
Ver la guía →
Agentes de IA

goclaw

OpenClaw rebuilt in Go — multi-tenant AI agent teams with 5-layer security in a single binary.

★ 3,6k estrellasCC BY-NC 4.0
Ver la guía →
Agentes de IA

lime

Full-stack AI agent for coding, research, content, and multi-agent workflows.

★ 1,5k estrellasCustom
Ver la guía →
Agentes de IA

Swarm

Swarm — OpenAI's experimental framework for lightweight multi-agent orchestration with agents and handoffs, MIT

★ 22,1k estrellasMIT
Ver la guía →
// alternativas

Alternativas en Agentes de IA

Los demás proyectos de Agentes de IA con más estrellas del catálogo de LibreAI: la lista realista de alternativas si Security Audit Skill no es lo que buscas.

  • openclaw★ 391,6k

    A personal AI that really does things — on any OS, with your data staying yours.

  • Skills For Real Engineers★ 322,7k

    Matt Pocock's engineering skills for coding agents — grill the spec first, then build it right, MIT

  • superpowers★ 294,9k

    An agentic skills framework & software development methodology that works.

  • ECC★ 275,2k

    Agent harness optimization system with skills, memory and security for Claude Code, Codex and more

  • hermes-agent★ 252,0k

    The agent that grows with you — a Python framework for adaptive AI agents.

// más en agents

Más en Agentes de IA

Proyectos de Agentes de IA cercanos en el catálogo, los más próximos en estrellas a Security Audit Skill: sigue explorando a ambos lados.

  • docling★ 68,5k

    Convert PDFs, DOCX, and HTML into clean, AI-ready documents.

  • mem0★ 66,8k

    Drop-in persistent memory infrastructure for AI agents and apps.

  • openinterpreter★ 68,5k

    A coding agent tuned for open models like Kimi, GLM, DeepSeek, and Qwen.

  • strix★ 66,3k

    Open-source AI penetration testing tool to find and fix your app’s vulnerabilities.

  • MetaGPT★ 70,7k

    🌟 The Multi-Agent Framework: First AI Software Company, Towards Natural Language Programming

  • Autogen★ 61,3k

    A programming framework for agentic AI

Consigue Security Audit Skill

Gratis y abierto: consigue el código fuente, dale una estrella al repositorio y empieza a crear.