mcp-maker

Pass

Audited by Gen Agent Trust Hub on Apr 5, 2026

Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted user specifications to generate source code, creating a surface for indirect prompt injection.
  • Ingestion points: User-provided names, descriptions, and logic for MCP tools collected in Step 2 of SKILL.md.
  • Boundary markers: The skill uses structured templates and explicit rules to treat user input as data rather than instructions, reducing the risk of logic overrides during generation.
  • Capability inventory: Capability to generate and write .py and .ts source files.
  • Sanitization: Enforces schema validation using Pydantic (Python) and Zod (TypeScript), implements path traversal prevention using realpath and prefix checks, and mandates an automated security audit of the output file.
  • [EXTERNAL_DOWNLOADS]: The skill references standard libraries and SDKs for MCP development, including 'claude-agent-sdk', 'fastmcp', and '@modelcontextprotocol/sdk'. These are well-known packages within the AI tooling ecosystem and are used as intended for server scaffolding.
  • [COMMAND_EXECUTION]: The skill documentation and internal rules explicitly prohibit the use of dangerous execution patterns, such as 'shell=True', 'exec()', or 'eval()', in the code it generates.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 5, 2026, 08:06 PM
Security Audit — agent-trust-hub — mcp-maker