lighthouse-bridge

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md requires parsing Lighthouse CI report artifacts and extracting audit descriptions/help text and affected element snippets from those reports at runtime, which are outsider-authored free text produced by CI runs (e.g., from audited URLs/pages) and then fed into the agent’s LLM context.