landing-page-design

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

• Third-party content exposure detected (low risk: 0.10). The skill documentation shows runtime paths like belt app run tavily/search-assistant and belt app run exa/answer, which indicates the agent actively searches/fetches outsider-authored web content based on user-supplied queries.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly relies on the inference.sh CLI (https://inference.sh) and runs remote apps via the belt CLI (e.g., falai/flux-dev-lora, tavily/search-assistant) which at runtime invoke https://inference.sh to execute remote model/app code, so external content controls execution.