tanstack-ai-core-media-generation

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

• Third-party content exposure detected (medium risk: 0.30). The runtime path ingests outsider-authored free text only via the request body fields (e.g., prompt, text, audio-derived transcription inputs) posted to the skill’s generation endpoints, with no provider-side filtering or required pre-selection of specific items.