mcp-builder
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructional content and reference implementation code for MCP server development.
- [EXTERNAL_DOWNLOADS]: Fetches protocol documentation and SDK README files from the official Model Context Protocol website and GitHub organization.
- [COMMAND_EXECUTION]: The evaluation harness script is designed to run local MCP servers as subprocesses using standard input/output, which is the intended functional use case for testing MCP integrations.
- [INDIRECT_PROMPT_INJECTION]: The evaluation script processes test cases from XML files and interpolates them into agent prompts.
- Ingestion points:
scripts/evaluation.pyreads user-provided XML evaluation files. - Boundary markers: XML tags delimit questions and answers.
- Capability inventory: The script executes tools provided by the connected MCP server.
- Sanitization: The script performs no sanitization on the content retrieved from the evaluation XML, but the tool is intended for controlled development use.
Audit Metadata