VibeCheck

Security scanner for vibe-coded apps. Catch what Copilot ships and Snyk overcharges for.

pip install vibecheck-ai
vibecheck scan .

What it catches

• Hardcoded secrets — AWS keys, API tokens, database URLs, private keys (16 patterns)
• SQL injection — f-strings and string concatenation in execute() calls
• Command injection — eval(), exec(), os.system(), subprocess with shell=True
• XSS vectors — dangerouslySetInnerHTML, document.write(), eval() in JS
• Missing auth — FastAPI/Express routes without authentication middleware
• Insecure config — DEBUG=True, CORS wildcard, insecure cookies
• Exposed .env — real secrets in .env files, missing .gitignore entries

Usage

# Scan current directory
vibecheck scan .

# Scan specific path
vibecheck scan ./src

# JSON output (for CI/CD)
vibecheck scan . --format json

# Only show critical and high
vibecheck scan . --severity high

# Quiet mode (summary only)
vibecheck scan . --quiet

Exit codes

• 0 — No critical or high findings
• 1 — Critical or high findings detected
• 2 — Scanner error

Supported languages

• Python (.py)
• JavaScript (.js, .jsx)
• TypeScript (.ts, .tsx)
• Environment files (.env)
• Config files (.json, .yaml, .toml)

Ignore patterns

Create a .vibeignore file in your project root:

# Skip test fixtures
tests/fixtures/*
# Skip generated code
generated/*

GitHub Action

Add to .github/workflows/vibecheck.yml:

name: VibeCheck
on: [push, pull_request]
jobs:
  scan:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      - uses: actions/setup-python@v5
        with:
          python-version: '3.12'
      - run: pip install vibecheck-ai
      - run: vibecheck scan .

For GitHub Code Scanning integration (findings appear inline on PRs):

      - run: vibecheck scan . --format sarif > vibecheck.sarif
      - uses: github/codeql-action/upload-sarif@v3
        with:
          sarif_file: vibecheck.sarif

Why VibeCheck?

53% of teams that shipped AI-generated code discovered security issues that passed review. The vibe coding era needs a security scanner that's:

• Fast — scans in milliseconds, not minutes
• Offline — no API calls, no code leaving your machine
• Free — open source, zero dependencies
• Focused — catches real issues, not style nits

License

MIT