monica

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

• Third-party content exposure detected (high risk: 0.75). The required runtime workflow calls Monica’s REST API and then JSON-decodes and prints the returned fields (e.g., request() → json.loads(raw) → out()), so any outsider-authored free text contained in Monica resources (like contact notes/bodies) can be ingested into the agent’s LLM context via the tool’s stdout.