interview-me

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

• Third-party content exposure detected (high risk: 0.80). The skill's mandatory Post-Flight Verification (in SKILL.md) instructs spawning a "claim-verifier" subagent and handing it external source pointers (DOIs, arXiv links, public/master_supporting_docs PDFs), which requires fetching and interpreting public third‑party documents whose content can change the spec (rewriting paragraphs or flagging citations).