mcp-validate

Validates MCP server tool definitions against the official Model Context Protocol specification by parsing tool metadata (name, description, input schema) and checking structural conformance to the spec's JSON Schema requirements. Uses schema introspection to ensure tools declare proper parameter types, required fields, and nested object structures before deployment.

Validates tool naming conventions and description quality by checking that tool names follow MCP naming rules (alphanumeric, underscores, hyphens), descriptions are present and sufficiently detailed, and metadata is LLM-readable. Performs pattern matching and length validation to ensure tools are discoverable and understandable by language models.

Validates that tool input schemas include proper documentation for all parameters by checking for descriptions in schema properties, ensuring required fields are marked, and verifying type definitions are complete. Inspects the JSON Schema structure recursively to catch undocumented nested properties and missing type constraints that would confuse LLMs.

Evaluates whether tool definitions are optimized for language model understanding by analyzing description clarity, parameter documentation, schema completeness, and naming conventions. Produces a readiness score or report indicating whether the tool definition will be effectively understood and used by Claude, GPT, or other LLMs when exposed through MCP.

Validates multiple tool definitions in a single operation and generates a comprehensive report showing which tools pass/fail validation, what errors were found, and which tools need remediation. Processes tool definitions from an MCP server registry or tool collection and produces structured output suitable for CI/CD pipelines or developer dashboards.