mcp-builder

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill utilizes FastMCP.from_openapi to generate tool schemas and logic dynamically at runtime from OpenAPI specifications, as demonstrated in assets/openapi-integration.py.
  • [INDIRECT_PROMPT_INJECTION]: The server templates in assets/tools-examples.py and assets/self-contained-server.py define tools that process user input without sanitization. Ingestion points: Parameters in search and data processing tools. Boundary markers: Absent in provided templates. Capability inventory: Network access via httpx, file system writes in CLI templates, and shell execution via mcp.run. Sanitization: Absent or basic string processing.
  • [COMMAND_EXECUTION]: The skill provides commands for the agent to install dependencies, compile code for syntax checks, and manage server processes.
  • [EXTERNAL_DOWNLOADS]: The skill initiates downloads of the FastMCP framework and its associated libraries from PyPI.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 07:27 AM
Security Audit — agent-trust-hub — mcp-builder