comfyui-video-production

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

• Third-party content exposure detected (high risk: 0.80). The skill's evolution/monitoring guidance (references/evolution.md "Monitoring Sources") explicitly instructs checking public sites (HuggingFace, GitHub repos, arXiv, Reddit, ComfyUI Discord, Civitai) and to evaluate new model releases from those untrusted, user-generated sources—content the agent would read and use to update model-selection and pipeline behavior—creating a clear avenue for indirect prompt injection.