web-fetch

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly provides HTTP fetch tools that accept arbitrary URLs (see "Available MCP Tools" and the tool parameters with "url") and the "Common Workflows" show fetching external endpoints and processing webhook payloads, so untrusted public web content and user-generated webhook data can be ingested and used to drive subsequent logic.

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

• Potentially malicious external URL detected (high risk: 0.90). The MCP configuration invokes npx to fetch and run the remote package @modelcontextprotocol/server-fetch at runtime ("npx -y @modelcontextprotocol/server-fetch"), which downloads and executes remote code that the skill depends on.