modular-mcp

🥈Silver

Modular MCP is a proxy server for managing large tool collections across multiple Model Context Protocol (MCP) servers. It groups tools and loads schemas on-demand, improving efficiency for operations teams using Claude agents.

5020Updated 1mo ago

Intermediate30min to implementautomation

Saves ~240 min per use

Quick InstallView Source

git clone https://github.com/d-kimuson/modular-mcp.git

Works with:

Claude

Overview

About This Skill

How to Use

[{"step":"Identify the MCP server and tool you need. Check the modular-mcp configuration to confirm the server is registered and the tool schema is available.","tip":"Use `list_servers` and `list_tools` commands in modular-mcp to verify availability before executing tasks."},{"step":"Construct your task with the correct tool name and parameters. Use the modular-mcp proxy to send the request, which will handle schema loading and tool execution.","tip":"If the tool requires complex parameters (e.g., nested JSON), validate the schema first using `get_tool_schema` to avoid errors."},{"step":"Execute the task via the modular-mcp proxy. Monitor the response for status, execution time, and any error messages.","tip":"For long-running tasks, check the `execution_time_ms` field to identify performance bottlenecks. Use this data to optimize future requests."},{"step":"Process the result or handle errors. If the tool fails, check the error message for clues (e.g., missing parameters, authentication issues).","tip":"For persistent issues, restart the modular-mcp server or check its logs for schema loading errors."},{"step":"Log the task and result for audit purposes. Use the metadata (e.g., file size, last modified) to track changes over time.","tip":"Integrate modular-mcp with your existing logging system to correlate tool usage with operational metrics."}]

Setup & Installation

Quick Install

No install command available. Check the GitHub repository for manual installation instructions.

Alternative Install (Git Clone)

git clone https://github.com/d-kimuson/modular-mcp

Requirements

Claude Code or compatible AI agent
Works with: Claude

Quick Start Guide

Install the Skill

Copy the install command above and run it in your terminal.

Open Your AI Agent

Launch Claude Code, Cursor, or your preferred AI coding agent.

Try It Out

Use the prompt template or examples below to test the skill.

Customize

Adapt the skill to your specific use case and workflow.

Usage Examples

Prompt Template

Use the modular-mcp proxy server to interact with [TOOL_NAME] from the [SERVER_NAME] MCP server. Execute the following task: [TASK_DESCRIPTION]. Ensure you reference the correct tool schema and handle any required parameters dynamically. If the tool requires authentication or context, assume it's already configured in the MCP server.

Example Output

```json
{
  "tool_used": "filesystem_read",
  "server": "local-fs-server",
  "parameters": {
    "path": "/projects/2024-q4/reports/sales_summary.md",
    "max_size": 1048576
  },
  "result": {
    "content": "# Q4 2024 Sales Report\n\n**Total Revenue**: $2.4M (↑15% YoY)\n**Top Product**: Widget Pro (42% of revenue)\n**Regional Performance**:\n- North America: $1.1M\n- EMEA: $850K\n- APAC: $450K\n\n**Key Risks**: Supply chain delays in APAC\n\nGenerated: 2024-10-15",
    "size": 1245,
    "last_modified": "2024-10-14T23:59:59Z"
  },
  "execution_time_ms": 124,
  "status": "success"
}
```

The modular-mcp proxy successfully retrieved the sales summary report from the local filesystem server. The tool schema was loaded on-demand, and the file was read within acceptable limits (1MB max size). The response includes metadata like file size and modification date, which can be used for validation. The execution time was minimal, demonstrating the efficiency of the proxy server's schema management.