loom-transcript

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

• Third-party content exposure detected (high risk: 1.00). The skill fetches and displays transcripts from arbitrary public Loom videos by calling https://www.loom.com/graphql and downloading the returned source_url and captions_source_url (VTT/JSON), exposing the agent to untrusted, user-generated content that it must read and present.