test-health

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime workflow executes scripts/inspect_test_health.py, which recursively reads and parses text files (e.g., manifests, CI configs, and sampled test files) from the repository root provided at --root, so outsider-authored repository contents can flow into the LLM context via the script’s collected file text/evidence.