mcp-builder
Warn
Audited by Snyk on Aug 11, 2026
Risk Level: MEDIUM
Full Analysis
MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).
- Third-party content exposure detected (medium risk: 0.30). The runtime evaluation harness reads outsider-authored free text from an evaluation XML file (
<question>strings) and passes it directly to the LLM as the user message inagent_loop, potentially triggering prompt injection via that text.
MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).
- Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs the agent/developer to WebFetch and load external documents at runtime — notably https://modelcontextprotocol.io/llms-full.txt (and the raw GitHub SDK READMEs at https://raw.githubusercontent.com/modelcontextprotocol/python-sdk/main/README.md and https://raw.githubusercontent.com/modelcontextprotocol/typescript-sdk/main/README.md) — which would supply external instructions/specifications that can directly control prompts or agent behavior.
Issues (2)
W011
MEDIUMThird-party content exposure detected (indirect prompt injection risk).
W012
MEDIUMUnverifiable external dependency detected (runtime URL that controls agent).
Audit Metadata