Inspector

Provides a debugging and testing surface for MCP tools, including an inspector, test client, and architectural tooling.

typescript

0

GitHub Stars

typescript

Language

7 months ago

First Indexed

3 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "mcp-mirror-stevennevins_architect-mcp-server": {
      "command": "npx",
      "args": [
        "@modelcontextprotocol/inspector",
        "node",
        "dist/index.js"
      ]
    }
  }
}

You run an MCP server to expose and test modular AI capabilities within your workspace. This server lets you interact with tools, run tests, and inspect tool behavior with a built-in inspector, helping you validate tool integrations and development flows before connecting to clients.

How to use

To use your MCP server with a client, start by launching the inspector for visual debugging and then run your tools through the MCP interface. The MCP server coordinates tool calls, tracks conversation context, and reports results in a structured way so your client can integrate them into workflows.

How to install

Prerequisites include having Node.js and a package manager available on your system. You also need the LLM CLI available in your PATH for architectural interactions.

Install the LLM CLI using your preferred package manager.

Verify the CLI is accessible by checking its version.

Install project dependencies for the MCP server by running the install command.

Start the development server for hot reload during development.

Build the project when you are ready to produce production assets.

Run tests to validate your tooling and integration points.

Start the production server to serve your MCP endpoints.

Additional sections

Architect Tool: This tool provides an interface to interact with the LLM CLI for architectural design feedback. It maintains conversation context, handles command execution through the LLM CLI, and offers error handling and logging. It supports both new conversations and continued discussions.

Testing: The template includes a built-in TestClient for local testing and the MCP Inspector for visual debugging.

MCP Inspector: Use the inspector to visually debug your MCP server, test tool calls, monitor requests and responses, inspect tool metadata, and interactively test endpoints.

Local testing with Cursor: Build and link the package, verify the binary, and add your server to Cursor to test it within that environment. You can also rebuild and relink after code changes and unlink when you are done.

Available tools

Architect tool

Interacts with the LLM CLI to design and review architectural aspects of your MCP setup. It preserves conversation context, executes commands via the LLM CLI, and handles errors and logging for both new and continued conversations.

TestClient

Provides a programmatic way to exercise MCP tools in tests. It validates tool calls and ensures responses contain expected structures.

MCP Inspector

Web-based debugging interface that shows available tools, their schemas, and supports interactive testing of tool calls and payload inspection.