MCP Config Validator
Validate MCP server configurations against the specification. Catch errors before deployment.
How to Use This Tool
- Paste your MCP server configuration JSON into the input area, or click 'Load Sample' to see a valid example.
- Click 'Validate' to run a comprehensive check against the MCP specification.
- Review the results: green checks for passed validations, red errors for issues that need fixing, and yellow warnings for best practice suggestions.
- If errors are found, fix them in your config and re-validate until all checks pass.
- The 'Discovered Capabilities' section shows a visual preview of all tools, resources, and prompts defined in your config.
Validating MCP Server Configurations
An MCP (Model Context Protocol) server configuration defines the capabilities your server exposes to AI clients. A valid configuration must include at least one of: tools (functions the AI can call), resources (data the AI can access), or prompts (template prompts the AI can use). Each type has specific schema requirements.
Common configuration mistakes include: missing required fields like tool descriptions, invalid parameter types in input schemas, duplicate tool or resource names, and incorrectly formatted resource URIs. These errors can cause AI clients to fail silently or produce unexpected behavior when interacting with your server.
This validator checks your configuration against the MCP specification and also provides best practice recommendations. For example, tool descriptions under 20 characters may not give AI clients enough context to use the tool correctly, and missing 'required' arrays in input schemas can lead to the AI omitting important parameters.
Use this validator during development to catch issues early. Run it before deploying your MCP server to production, and re-validate after any configuration changes. Combined with the MCP Manifest Generator, you have a complete workflow for creating and validating MCP server configs.
Last updated: February 2026
FAQ
What does this validator check?
It validates JSON syntax, required fields (tools, resources, prompts), tool naming conventions, input schema structure, parameter types, duplicate names, and best practices like description length and required array presence.
What is a valid MCP configuration?
A valid MCP config is a JSON object with at least one of: tools (array of tool definitions), resources (array of resource URIs), or prompts (array of prompt templates). Each tool must have a name, description, and inputSchema with type 'object'.
Does this validate against the latest MCP spec?
Yes, this validator checks against the MCP specification as of early 2026. It validates the core schema requirements and also provides best practice warnings based on community conventions.
How do I fix 'invalid tool name' errors?
MCP tool names should be lowercase with hyphens or underscores (e.g., 'search-database' or 'get_user_info'). Avoid spaces, uppercase letters, or special characters. This convention helps AI clients consistently reference and call your tools.
What if my config has no errors but the AI client still fails?
Schema validation is necessary but not sufficient. Your MCP server implementation must match the config exactly — the actual function signatures, parameter handling, and response formats must align with what the config declares. Test with a real AI client after validation.