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 recommends arbitrary external image URLs via the input_image / input_image_2 ... parameters (e.g., "input_image": "https://example.com/photo.jpg") which the API "fetches URLs automatically" and the model consumes for image editing, so it ingests untrusted third‑party content as part of its workflow.