bfl-api

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly accepts and instructs using arbitrary public image URLs (see "Image Input for Editing" in SKILL.md and the endpoints/code examples like references/endpoints.md and code-examples/*) which the API/client will fetch and ingest — untrusted third-party content that the model is expected to interpret and that can materially change generation outputs.