follow-up-after-call

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

• Third-party content exposure detected (high risk: 0.75). SKILL.md requires “Grounded call evidence” and explicitly allows fetching/using meeting transcripts/notes and other customer-facing thread content (outsider-authored text) as LLM-readable input for synthesis, so the runtime workflow can ingest free-form outsider text via the call-evidence retrieval path.