customer-persona

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

• Third-party content exposure detected (high risk: 0.85). The required workflow uses runtime web-retrieval tools like belt app run tavily/search-assistant and belt app run exa/search in Step 1 (Research), which fetch outsider-authored public web content and feed the extracted text into the agent/LLM context.

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 platform via the belt CLI (https://inference.sh) to run external apps (e.g., tavily/search-assistant, falai/flux-dev-lora) at runtime, which executes remote code and provides/influences prompts and outputs, making it a required external runtime dependency.