lighthouse
Warn
Audited by Snyk on Feb 21, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and audits arbitrary URLs (e.g., TARGET_URL set from $ARGUMENTS and examples like "/lighthouse https://your-production-site.com") using curl and the run_lighthouse_audit/lighthouse command, then parses the resulting REPORT_JSON and generates fixes based on those audit results—so untrusted public web pages can be ingested and materially influence actions.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.70). The skill installs and then runs the Lighthouse CLI at runtime via "npm install -g lighthouse", which fetches and will execute remote code from the npm registry (e.g. https://registry.npmjs.org/lighthouse or https://www.npmjs.com/package/lighthouse), and the skill relies on that dependency to perform audits.
Audit Metadata