mcp-management

Discover, analyze, and execute MCP capabilities across multiple servers.

Updated Feb 27, 2026
One-click install
npx skills add https://github.com/haidonglethqb/CloudSchool --skill mcp-management-haidonglethqb
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: mcp-management
Source: https://github.com/haidonglethqb/CloudSchool/tree/main/.claude/skills/mcp-management
Command: npx skills add https://github.com/haidonglethqb/CloudSchool --skill mcp-management-haidonglethqb

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires @modelcontextprotocol/sdk, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

MCP-management provides a unified way to discover and orchestrate Model Context Protocol servers to avoid context pollution and duplicate effort.

Core Features & Use Cases

  • Discover MCP capabilities across multiple servers with a single config
  • Analyze and select relevant tools/prompts/resources for tasks
  • Execute tools across servers and report results back to the main agent

Quick Start

Run a simple MCP discovery to list all tools from your configured servers.

Frequently Asked Questions about mcp-management

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I manage context pollution when integrating multiple MCP servers?

To manage context pollution across multiple MCP servers, use a subagent pattern for execution that isolates tool discovery and execution, reporting results back to the main agent without bloating the primary context window.

What is the best way to discover tools across multiple MCP servers?

The best way to discover tools across multiple MCP servers is using a unified discovery operation that reads a local configuration file to list available tools, prompts, and resources across all configured servers in a single query.

Do I need a specific config file to use the Model Context Protocol for multi-server orchestration?

Yes, multi-server MCP orchestration requires a local configuration file located at .claude/.mcp.json, which defines the servers available for discovery, analysis, and cross-server tool execution.

Can I execute tools across different MCP servers without loading everything into the main context?

Yes, you can execute tools across different MCP servers without loading everything into the main context by using a subagent execution pattern, where the subagent handles the tool call and returns only the final result.

How does cross-server tool selection work with the Model Context Protocol?

Cross-server tool selection works by analyzing capabilities listed from all configured MCP servers, matching them to the task, selecting the most relevant tools or prompts across servers, and executing them via a subagent.

Why does my MCP server discovery duplicate effort and waste context tokens?

MCP server discovery duplicates effort and wastes tokens when each server is queried individually within the main context; a unified discovery and subagent execution workflow prevents this by centralizing configuration and isolating execution.