The Agent Skills Directory

[SAFE]: The skill serves as a technical manual for developers building MCP integrations. The instructions are consistent with industry best practices for API integration and tool design.
[EXTERNAL_DOWNLOADS]: The skill directs the agent to fetch documentation from official sources, including modelcontextprotocol.io and github.com/modelcontextprotocol. These downloads are for protocol specification and SDK documentation, which are necessary for the skill's stated purpose.
[INDIRECT_PROMPT_INJECTION]: The skill demonstrates an attack surface by ingesting external data through WebFetch.
Ingestion points: Section 1.2 and 1.3 in SKILL.md fetch content from remote URLs to provide context for the developer.
Boundary markers: None explicitly defined for the documentation fetch, but the context is clearly restricted to technical reference.
Capability inventory: The skill defines a process for tool creation but does not contain active scripts that execute commands or perform network operations on its own.
Sanitization: The guide explicitly instructs users to implement strict input validation using Pydantic (Python) or Zod (TypeScript) in the servers they develop.
[COMMAND_EXECUTION]: The skill mentions standard development commands like python -m py_compile and npm run build for verifying user-written code. These are benign and intended for developer verification during the build process.

mcp-builder