Back to the catalog

snitch

Security audit plugin for Claude Code, Gemini CLI, Codex CLI, OpenCode, Antigravity, and Cursor

Open source Open in the app JSON README (API)

About

Security audit plugin for Claude Code, Gemini CLI, Codex CLI, OpenCode, Antigravity, and Cursor

Details

Kind
Plugins
Topic
AI, RAG & memory
Publisher
jf10r
Origin
gemini
Category
ferramentas
Version
5.2.0
Stars
2
Forks
1
Last push
2026-03-23T13:15:04Z
Repository state
ativo
License
MIT
Added
2026-08-30 14:13:39
Updated
2026-08-30 14:13:39
Origin id
jf10r/snitch

README

```
   _____ _   ____________________  __
  / ___// | / /  _/_  __/ ____/ / / /
  \__ \/  |/ // /  / / / /   / /_/ /
 ___/ / /|  // /  / / / /___/ __  /
/____/_/ |_/___/ /_/  \____/_/ /_/
```

Evidence-based security auditing for AI coding assistants. Zero false positives.

Compatible with [Claude Code](https://docs.anthropic.com/en/docs/claude-code) · [Gemini CLI](https://github.com/google-gemini/gemini-cli) · [Codex CLI](https://github.com/openai/codex) · [OpenCode](https://github.com/opencode-ai/opencode) · [Antigravity](https://github.com/neplextech/antigravity) · [Cursor](https://www.cursor.com/)

---

## Why Snitch

Traditional scanners flood you with hundreds of findings — half of them are `YOUR_API_KEY_HERE` in a comment. Snitch is different: every finding must be backed by real code evidence. No file read? No finding. Can't quote the exact line? No finding. Didn't check for a fix nearby? No finding.

Findings are tagged with **CWE**, **OWASP Top 10:2025**, and **CVSS 4.0** references for direct GRC ingestion.

---

## Installation

### Claude Code

```
/plugin marketplace add JF10R/Snitch
/plugin install snitch@JF10R-Snitch
```

### Gemini CLI

```bash
gemini extensions install https://github.com/JF10R/Snitch.git
```

### Codex CLI

```bash
git clone https://github.com/JF10R/Snitch.git
cp -r Snitch/agents/skills/snitch ~/.codex/skills/snitch
```

Per-project: use `.agents/skills/snitch` instead.

### OpenCode

```bash
git clone https://github.com/JF10R/Snitch.git
cp -r Snitch/skills/snitch ~/.config/opencode/skills/snitch
```

Per-project: `.opencode/skills/snitch`

### Antigravity

```bash
git clone https://github.com/JF10R/Snitch.git
cp -r Snitch/skills/snitch ~/.gemini/antigravity/skills/snitch
```

Per-project: `.agent/skills/snitch`

### Cursor

```bash
git clone https://github.com/JF10R/Snitch.git
cp -r Snitch/skills/snitch .cursor/skills/snitch
```

<details>
<summary>Updating / Uninstalling</summary>

| Platform | Update | Uninstall |
|----------|--------|-----------|
| Claude Code | `/plugin marketplace update JF10R-Snitch` | `/plugin uninstall snitch@JF10R-Snitch` |
| Gemini CLI | `gemini extensions update snitch` | `gemini extensions uninstall snitch` |
| Codex CLI | Re-clone and copy | `rm -rf ~/.codex/skills/snitch` |
| OpenCode | Re-clone and copy | `rm -rf ~/.config/opencode/skills/snitch` |
| Antigravity | Re-clone and copy | `rm -rf ~/.gemini/antigravity/skills/snitch` |
| Cursor | Re-clone and copy | `rm -rf .cursor/skills/snitch` |

</details>

---

## Usage

```
/snitch
```

Select categories from the interactive menu, or run directly:

```
/snitch --categories=1,2,3,13
/snitch --diff
```

**Quick Scan** detects your stack automatically (`package.json`, imports, config files) and only audits relevant categories.

**`--diff`** scans only staged/unstaged changes — ideal as a pre-commit check.

### After the scan

- **Fix one by one** — walk through each finding individually
- **Fix all** — auto-patch everything at once
- **Run another scan** — check additional categories
- **Done** — exit

---

## Categories

40 audit categories organized by domain:

### Application Security

| # | Category | Description |
|---|----------|-------------|
| 01 | SQL Injection | Parameterized queries, ORM misuse |
| 02 | XSS | Output encoding, DOM injection |
| 03 | Hardcoded Secrets | API keys, passwords, tokens in source |
| 04 | Authentication | Login flows, password handling, MFA |
| 05 | SSRF | Server-side request forgery |
| 07 | Rate Limiting | Brute-force protection, throttling |
| 08 | CORS | Cross-origin misconfiguration |
| 09 | Cryptography | Weak algorithms, key management |
| 10 | Dangerous Patterns | `eval()`, dynamic code execution |
| 28 | Authorization | Broken access control, IDOR |
| 29 | File Uploads | Validation, path traversal |
| 30 | Input Validation | ReDoS, injection vectors |
| 32 | Security Headers | CSP, HSTS, X-Frame-Options |
| 39 | Token Lifetimes | Session expiry, logout effectiveness |

### Services & Integrations

| # | Category | Description |
|---|----------|-------------|
| 06 | Supabase | Row-level security, exposed service keys |
| 13 | Stripe | API keys, webhook verification |
| 14 | Auth Providers | Clerk, Auth0, NextAuth configuration |
| 15 | AI APIs | Claude, OpenAI, Gemini key exposure |
| 16 | Email | SMTP credentials, spam abuse vectors |
| 17 | Database | Connection strings, query security |
| 18 | Redis | Authentication, exposed instances |
| 19 | SMS | Twilio tokens, message injection |

### Infrastructure

| # | Category | Description |
|---|----------|-------------|
| 11 | Cloud Providers | AWS, GCP, Azure, Vercel, Cloudflare |
| 12 | Data Leaks | Logs, error messages, debug output |
| 31 | CI/CD | Pipeline secrets, deployment security |
| 40 | Tunnels & DNS | ngrok, cloudflared, DNS configuration |

### Compliance

| # | Category | Description |
|---|----------|-------------|
| 20 | HIPAA | Protected health information |
| 21 | SOC 2 | Audit trails, access controls |
| 22 | PCI-DSS | Payment card data handling |
| 23 | GDPR | Data deletion, consent, EU requirements |
| 34 | FIPS 140-3 | Cryptographic module compliance |
| 35 | Governance | ISO 27001, FedRAMP, CMMC |
| 38 | Data Classification | Sensitivity labeling, handling policies |

### Performance & Maintenance

| # | Category | Description |
|---|----------|-------------|
| 24 | Memory Leaks | Event listeners, uncleaned resources |
| 25 | N+1 Queries | ORM batching, query optimization |
| 26 | Performance | Blocking I/O, CPU-bound operations |
| 27 | Dependencies | Known CVEs, outdated packages |
| 33 | Unused Dependencies | Dead code, bundle bloat |
| 36 | BCDR | Backup, disaster recovery |
| 37 | Monitoring | Observability, alerting gaps |

---

## How It Works

Snitch is a skill file (`SKILL.md`) — no runtime, no build step, no dependencies. It guides AI assistants through structured security analysis with built-in anti-hallucination rules:

1. **Read** the actual source files before reporting
2. **Verify** each finding with exact file path and line number
3. **Check context** — is there a fix nearby? Is this test code? Server or client?
4. **Prove it** — no evidence, no finding

---

## Contributing

See [CONTRIBUTING.md](.github/CONTRIBUTING.md) for guidelines.

## License

[MIT](LICENSE)

More