meetings-recap

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

• Third-party content exposure detected (high risk: 0.70). SKILL.md:7-14 describes fetching meeting files via the meetings-search skill and then using Read on the returned paths, which at runtime will ingest outsider-authored meeting/voice-memo text (e.g., recordings/transcripts/comments not authored by the operating user) into the LLM context for synthesis.