clearance

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 includes “Similar-marks search against what's connected (Solve Intelligence, CourtListener, Descrybe, or whatever MCP is available)” and then “Attribute every result to its source,” which implies it ingests outsider-authored free text from those external search results/pages into the LLM context for summarization and memo writing.