deep-research

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

• Third-party content exposure detected (high risk: 0.75). The skill’s “Deep Research” workflow requires gathering information from multiple sources (step 3) and synthesizing with citations, which at runtime typically involves fetching/reading outsider-authored web pages or documents; those readable texts would be ingested into the agent’s LLM context for summarization.