codex-imagegen

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

• Third-party content exposure detected (medium risk: 0.30). In scripts/generate.py, the outsider-provided prompt/--negative/--style/--aspect strings are embedded into the runtime instruction passed to codex exec (via build_prompt() → build_codex_instruction() → run_codex()), so the required workflow directly ingests free text supplied by the user.