add-model-descriptions

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

• Third-party content exposure detected (high risk: 0.90). This skill fetches model listings from the public HuggingFace router (WebFetch https://router.huggingface.co/v1/models) and explicitly instructs the agent to "search the web for [model] specifications", meaning it ingests and interprets untrusted public web content that could contain malicious or adversarial instructions.