mcp-builder
Warn
Audited by Snyk on Mar 6, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly instructs the agent to fetch and study public web pages (e.g., "Start with the sitemap at https://modelcontextprotocol.io/sitemap.xml" and "Use WebFetch to load https://raw.githubusercontent.com/.../README.md") as required steps in its workflow, meaning the agent ingests untrusted third-party content that can influence tool design and subsequent actions.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.90). The evaluation harness accepts a user-supplied MCP server URL (e.g., https://example.com/mcp) and at runtime calls the remote server (connection.list_tools) and passes the returned tools into the Anthropic call, so content fetched from that URL directly controls the agent's available tools and thus its prompts/behavior.
Audit Metadata