How do I check a specific server instead of all of them?

You can target a specific server by using the --server flag followed by the server name when running the doctor command.

Why should I care about duplicate tools?

Duplicate tool definitions across multiple MCP servers increase the size of the system prompt, wasting context window tokens and potentially confusing the AI model.

Does this skill require special permissions?

It requires the ability to execute the local Python diagnostic script included in the plugin directory to perform real connection tests.

Can MCP Doctor automatically fix my configuration?

For safety, it does not modify files directly. Instead, it provides specific, copy-pasteable fix commands to resolve issues like credential errors or broken connections.

What does MCP Doctor actually check?

It performs a full initialization and tool listing for each server, measures response time in milliseconds, and scans for missing credentials or duplicate tool schemas.

MCP Doctor

Name: MCP Doctor
Author: choam2426

bychoam2426

•

分析与监控

Diagnoses and optimizes Model Context Protocol (MCP) server health, connectivity, and tool redundancy.

MCP Doctor provides a comprehensive diagnostic toolkit for Claude Code, allowing users to verify MCP server health, measure latency, and identify configuration errors. By running real-time connection tests and scanning for missing credentials or duplicate tool definitions, it ensures your AI environment is stable and token-efficient. Whether you are troubleshooting a failing server or auditing your setup for performance, MCP Doctor offers actionable fix recommendations to streamline your development workflow and reduce context bloat.

主要功能

018 GitHub stars

02Automated detection of missing environment variables and placeholder credentials

03Real-time connection and initialization testing for configured MCP servers

04Identification of duplicate tool names across servers to prevent token waste

05Precise measurement of server response times to identify performance bottlenecks

06Actionable fix commands for resolving broken configurations and redundancies

使用场景

01Auditing your environment to remove redundant tool definitions and optimize token usage

02Troubleshooting why an MCP server is failing to load or timing out in Claude Code

03Verifying that all necessary environment variables are correctly passed to your MCP servers

主要功能

018 GitHub stars

02Automated detection of missing environment variables and placeholder credentials

03Real-time connection and initialization testing for configured MCP servers

04Identification of duplicate tool names across servers to prevent token waste

05Precise measurement of server response times to identify performance bottlenecks

06Actionable fix commands for resolving broken configurations and redundancies

使用场景

01Auditing your environment to remove redundant tool definitions and optimize token usage

02Troubleshooting why an MCP server is failing to load or timing out in Claude Code

03Verifying that all necessary environment variables are correctly passed to your MCP servers