The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/evaluation.py and scripts/connections.py files provide a harness for launching local processes (MCP servers) via subprocess. This is an intended feature that allows developers to test their own code in a local environment.
[EXTERNAL_DOWNLOADS]: The documentation provides instructions for the agent to fetch the Model Context Protocol specification and SDK documentation from the official modelcontextprotocol GitHub organization, which is a trusted source for this technology.
[INDIRECT_PROMPT_INJECTION]: The evaluation script in scripts/evaluation.py ingests data from external MCP tool outputs and XML-formatted questions. This creates a potential surface for indirect prompt injection if a tested server returns malicious instructions; however, this is a standard risk for a protocol testing harness and is handled by the sub-agent's prompt constraints.

mcp-builder