blink-granola

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

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text can enter the agent’s LLM context via the runtime tool outputs from Granola’s MCP—e.g., query_granola_meetings, get_meetings, and get_meeting_transcript return meeting notes/transcripts authored by other participants, which the agent then uses in its responses.