deep-research-article

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

• Third-party content exposure detected (high risk: 0.90). The skill's default workflow requires an "internet-backed" Step 1 — "Internet source sweep (required by default)" and references/web_research_protocol.md to fetch, read, and extract claims from public web sources/URLs into source_dossier.md and evidence_extract.md, and those extracted third‑party contents are used to build the claim→evidence ledger and drive drafting/next actions, creating a clear vector for indirect prompt injection.