legal-risk-analysis-goodlegal

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

• Third-party content exposure detected (medium risk: 0.65). The skill’s runtime workflow is primarily internal risk scoring, but it explicitly instructs using web_search to fetch doctrinal commentary from the public web and case_retrieve/case_search to pull full decision text, which can include outsider-authored free-form prose that is ingested into the LLM context (indirect prompt injection risk).