moai-cc-mcp-builder

Pass

Audited by Gen Agent Trust Hub on Mar 1, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface via its documentation retrieval process.\n
  • Ingestion points: External design patterns and best practices are ingested from the Context7 library using the mcp__context7__get-library-docs tool and WebFetch to guide the AI-powered code generation.\n
  • Boundary markers: The skill lacks explicit delimiters or specific system instructions to ignore potentially malicious instructions embedded within the fetched documentation or external data.\n
  • Capability inventory: The skill is authorized to use Bash, Write, and Edit tools, and it implements a subprocess execution wrapper in scripts/connections.py.\n
  • Sanitization: No evidence of sanitization or content validation exists for the remote data before it is processed by the LLM to generate or modify server code.\n- [COMMAND_EXECUTION]: The skill executes system commands to facilitate the testing and evaluation of MCP servers.\n
  • Evidence: The scripts/evaluation.py and scripts/connections.py scripts allow the agent to launch MCP servers via the stdio transport, which executes subprocesses using commands and arguments provided via CLI or agent tools.\n- [EXTERNAL_DOWNLOADS]: The skill fetches dynamic content from a remote vendor service.\n
  • Evidence: The orchestrator is designed to retrieve 'latest MCP patterns' and 'development intelligence' from the Context7 library, which serves as a remote repository for development standards.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 1, 2026, 01:06 AM