# Security Audit (Cloudflare)

> A Cloudflare skill for multi-phase security audits with verified findings

- Category: Agent skills (`skills`)
- Tags: `free`, `open-source`
- Official: https://github.com/cloudflare/security-audit-skill
- Wunderlust: https://wunderlust.xyz/r/cloudflare-security-audit
- Image: https://wunderlust.xyz/media/cloudflare-security-audit-poster.jpg
- Video preview: https://wunderlust.xyz/media/cloudflare-security-audit.webm
- Added: 2026-11-05

## What it is

A Cloudflare agent skill that runs a security audit in six phases, with each finding verified by an independent agent.

## What it’s for

Use it to ask an agent for a structured audit of a repository, with output in validatable JSON.

## Benefits

- Six phases: reconnaissance, coverage-driven hunting, validation, structured output, independent verification and report
- Classifies findings as confirmed, needs_validation or rejected
- Guides per attack class: memory, LLM apps, auth, supply chain, cloud and mobile

## Caveats

- MIT license, no pricing
- Requires an agent with parallel subagents, Node.js and an OS sandbox; without a sandbox, findings stay as needs_validation

## Similar on Wunderlust

- [Emil Kowalski Design Skill](https://wunderlust.xyz/r/emil-kowalski-skills.md): Emil Kowalski's skills for animation and interface design with agents
- [Design Motion Principles](https://wunderlust.xyz/r/design-motion-principles.md): A motion skill with modes for creating and auditing interface animations
- [Apple Design Skill](https://wunderlust.xyz/r/apple-design-skill.md): Agent skill with Apple-style interface design guidelines
- [Anatomy](https://wunderlust.xyz/r/anatomy.md): Skill that creates interactive isometric machine illustrations to explain technical ideas
- [Impeccable](https://wunderlust.xyz/r/impeccable.md): Design vocabulary and commands for agents to review and refine interfaces
- [No AI Slop](https://wunderlust.xyz/r/no-ai-slop.md): Skill that edits text to remove telltale habits of AI writing