claim-verification

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

• Third-party content exposure detected (high risk: 0.75). Outsider free text can enter via Pass 2’s runtime network fallback: the skill invokes superpower-writing:research-lookup / citation-management to fetch abstracts from Crossref/PubMed (public web content), then performs an LLM semantic match over that retrieved abstract text in the agent context.