newegg-pc-builder

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

• Third-party content exposure detected (high risk: 0.80). The skill dynamically queries the external Newegg MCP endpoint (https://apis.newegg.com/ex-mcp/endpoint/pcbuilder) and its SKILL.md plus scripts/mcp_client.py explicitly require the LLM to read the runtime tool descriptions, parameter schemas (list_tools) and returned content and then pick tools/parameters and act on responses, meaning untrusted third‑party content from that API can materially influence agent behavior.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill makes runtime requests to the MCP endpoint https://apis.newegg.com/ex-mcp/endpoint/pcbuilder which returns tool descriptions and input schemas that directly drive the model's prompt-generation and decision logic and are required for the skill to function.