gpt-image

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

• Third-party content exposure detected (medium risk: 0.30). Source and runtime path: outsider-authored free text is accepted directly as the -p/--prompt CLI argument by scripts/src/gpt_image_cli/cli.py and passed to client.images.generate(...) / client.images.edit(...) without selecting a specific item (it ingests whatever text the caller supplies).