# Security Audit

> Run an OWASP-style security audit on the current diff, the working tree, or a specified path. Flags injection, auth flaws, secrets, insecure deserialization, and misconfigurations.

- **Type:** Skill
- **Install:** `agentstack add skill-kasimmj-claude-skills-mega-security-audit`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [kasimmj](https://agentstack.voostack.com/s/kasimmj)
- **Installs:** 0
- **Category:** [AI & ML](https://agentstack.voostack.com/c/ai-and-ml)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [kasimmj](https://github.com/kasimmj)
- **Source:** https://github.com/kasimmj/claude-skills-mega/tree/main/skills/security-audit
- **Website:** https://kasimmj.github.io/claude-skills-mega/

## Install

```sh
agentstack add skill-kasimmj-claude-skills-mega-security-audit
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

# Security Audit Skill

You are performing a **security audit**. Output a structured report — not just prose.

## Step 1 — Define the scope

Ask the user (or infer) the scope:
- **Diff** — current uncommitted changes (default if there are uncommitted changes)
- **Branch** — `git diff main...HEAD`
- **Repo** — full working tree
- **Path** — a specific directory or file

## Step 2 — Run the audit

Walk through the OWASP Top 10 (2021) categories systematically:

1. **A01 Broken Access Control** — missing authz checks, IDOR, path traversal
2. **A02 Cryptographic Failures** — weak hashing (MD5, SHA1), hardcoded keys, missing TLS
3. **A03 Injection** — SQL, NoSQL, OS command, LDAP, XSS — search for unparameterized string interpolation in queries, untrusted input in `exec`/`eval`
4. **A04 Insecure Design** — race conditions in critical flows (payments, account changes)
5. **A05 Security Misconfiguration** — debug=true in prod, default creds, exposed admin panels, permissive CORS
6. **A06 Vulnerable Components** — outdated deps with known CVEs (cross-reference package.json/requirements.txt/go.mod)
7. **A07 Identification & Authentication Failures** — weak password policy, missing rate-limit on login, JWT with `alg=none`, sessions without timeout
8. **A08 Software & Data Integrity Failures** — unsigned updates, insecure deserialization (pickle, YAML.load, Java ObjectInputStream)
9. **A09 Logging & Monitoring Failures** — sensitive data in logs, missing audit trails
10. **A10 SSRF** — unvalidated URL fetches that could hit internal services (169.254.169.254, localhost)

## Step 3 — Look for secrets

Search for committed secrets using these patterns:
- `aws_secret`, `AKIA[0-9A-Z]{16}`, `ghp_[A-Za-z0-9]{36}`, `sk-[A-Za-z0-9]{32,}`
- `password\s*=\s*['"]`, `api[_-]?key\s*=\s*['"]`
- `.env`, `.pem`, `.key` files that should be gitignored

## Step 4 — Output the report

```
SECURITY AUDIT REPORT
─────────────────────
Scope:    
Files:    

🔴 CRITICAL (N)
  -  —  — 

🟠 HIGH (N)
  -  —  — 

🟡 MEDIUM (N)
  -  —  — 

🟢 LOW (N)
  -  —  — 

✓ PASSED CHECKS
  - : clean
```

## When NOT to use

- Repos with no source code (docs-only)
- When the user wants a code review (use `code-review` instead)
- When the user wants a license scan (use `license-check`)

## Failure modes

- ⚠️ A clean report doesn't mean the code is secure — it means *these checks* passed.
- ⚠️ Always recommend a real third-party scanner (Semgrep, Snyk, Trivy) for production releases.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [kasimmj](https://github.com/kasimmj)
- **Source:** [kasimmj/claude-skills-mega](https://github.com/kasimmj/claude-skills-mega)
- **License:** MIT
- **Homepage:** https://kasimmj.github.io/claude-skills-mega/

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** yes
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-kasimmj-claude-skills-mega-security-audit
- Seller: https://agentstack.voostack.com/s/kasimmj
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.