2547 Github Actions 172d2106

GitHub Actions Integration

Skill Scanner provides a reusable workflow you can call from any repository to scan Agent Skills on every push or pull request. Results can be uploaded to GitHub Code Scanning for inline annotations.

Quick Start (Static Analysis Only, No Keys Required)

Add this file to your repository at .github/workflows/scan-skills.yml:

name: Scan Skills

on:
  push:
    paths: [".cursor/skills/**"]
  pull_request:
    paths: [".cursor/skills/**"]

jobs:
  scan:
    uses: cisco-ai-defense/skill-scanner/.github/workflows/scan-skills.yml@main
    with:
      skill_path: .cursor/skills
    permissions:
      security-events: write
      contents: read

This will:

1. Install cisco-ai-skill-scanner from PyPI on a fresh runner
2. Run skill-scanner scan-all .cursor/skills --format sarif --recursive --check-overlap
3. Upload SARIF results to GitHub Code Scanning (findings appear as annotations on PRs)
4. Fail the workflow if any findings at or above HIGH severity are detected (configurable via fail_on_severity)

Reusable Workflow Inputs

Secrets

All secrets are optional and only needed for advanced analysis features.

To configure secrets, go to your repository's Settings > Secrets and variables > Actions and add them there. They are never exposed in logs.

Configuration Tiers

Tier 1: Static Analysis (No Keys)

Zero-config static scanning with YARA rules, behavioral analysis, and SARIF upload:

jobs:
  scan:
    uses: cisco-ai-defense/skill-scanner/.github/workflows/scan-skills.yml@main
    with:
      skill_path: .cursor/skills

Tier 2: Static + LLM (One Key)

Add LLM-powered semantic analysis for deeper threat detection:

jobs:
  scan:
    uses: cisco-ai-defense/skill-scanner/.github/workflows/scan-skills.yml@main
    with:
      skill_path: .cursor/skills
      use_llm: true
      llm_model: gpt-4o
    secrets:
      llm_api_key: ${{ secrets.SKILL_SCANNER_LLM_API_KEY }}

Tier 3: Full Stack (All Keys)

Enable every analyzer including VirusTotal binary scanning:

jobs:
  scan:
    uses: cisco-ai-defense/skill-scanner/.github/workflows/scan-skills.yml@main
    with:
      skill_path: .cursor/skills
      use_llm: true
      use_behavioral: true
      policy: strict
    secrets:
      llm_api_key: ${{ secrets.SKILL_SCANNER_LLM_API_KEY }}
      virustotal_api_key: ${{ secrets.VIRUSTOTAL_API_KEY }}

Branch Protection

To block PRs with security findings:

1. Go to Settings > Branches > Branch protection rules
2. Enable Require status checks to pass before merging
3. Search for and select the Skill Scanner check
4. Save changes

Now any PR that touches skill files must pass the security scan before it can be merged.

Self-Hosted Workflow (Copy-Paste)

If you prefer not to use the reusable workflow, copy this standalone workflow into your repo:

name: Scan Skills

on:
  push:
    paths: [".cursor/skills/**"]
  pull_request:
    paths: [".cursor/skills/**"]

jobs:
  scan:
    runs-on: ubuntu-latest
    permissions:
      security-events: write
      contents: read
    steps:
      - uses: actions/checkout@v4

      - uses: actions/setup-python@v5
        with:
          python-version: "3.12"

      - run: pip install cisco-ai-skill-scanner

      - name: Scan skills
        run: |
          skill-scanner scan-all .cursor/skills \
            --format sarif \
            --output results.sarif \
            --recursive \
            --check-overlap \
            --fail-on-severity high

      - name: Upload SARIF
        if: always()
        uses: github/codeql-action/upload-sarif@v4
        with:
          sarif_file: results.sarif