mcp-developer
Pass
Audited by Gen Agent Trust Hub on Sep 4, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill provides instructions for installing official development SDKs and tools via standard package managers (npm and pip).
- [REMOTE_CODE_EXECUTION]: The documentation includes the use of
npxto execute scaffolding and debugging tools for the Model Context Protocol. - [COMMAND_EXECUTION]: Reference implementations include examples of tools that interact with the local environment, such as executing git commands or database queries, following standard MCP integration patterns.
- [INDIRECT_PROMPT_INJECTION]: The skill describes building systems that ingest untrusted data (Ingestion points:
references/tools.md,references/resources.md). It implements boundary markers through defined schemas and provides a capability inventory including subprocess and database execution. Sanitization is strictly enforced through Zod/Pydantic validation and path resolution logic (Sanitization:references/python-sdk.md,references/resources.md).
Audit Metadata