secrets-gitleaks

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The workflow template reads JSON/SARIF findings and (on failure) posts derived text to GitHub/Slack, where the underlying finding content ultimately comes from the repository’s outsider-authored commit contents that triggered the scan (e.g., gitleaks-report.json / gitleaks.sarif generated from the checked-out repo).

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill contains runtime commands that fetch and execute remote code — e.g., cloning and then running scripts from https://github.com/org/repo.git and pulling/running the Docker image zricethezav/gitleaks:latest, plus installing pre-commit hooks from https://github.com/gitleaks/gitleaks and https://github.com/pre-commit/pre-commit-hooks — which are external artifacts executed at runtime.