home / mcp / agile luminary mcp server

Agile Luminary MCP Server

Provides access to Agile Luminary data via MCP: current work, related documents, and past work.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "agileluminary-mcp-agile-luminary": {
      "command": "node",
      "args": [
        "server.js"
      ],
      "env": {
        "LUMINARY_API_KEY": "YOUR_API_KEY"
      }
    }
  }
}

You run an MCP server that connects AI clients to the Agile Luminary project management system. It lets your AI assistant fetch current work, related documents, and past work from Agile Luminary directly within conversations, enabling real-time project context and smarter task planning.

How to use

You interact with the Agile Luminary MCP Server through your MCP-enabled AI client. Once the server is running, your AI can ask for current assignments, search for related documents, or retrieve past work related to a user story. Use natural language prompts such as: “Show me my current work,” “Find documents about authentication,” or “What past work relates to the user login feature?” The server handles authentication locally and forwards requests to the Agile Luminary API, returning structured data for your AI client to present.

How to install

Prerequisites you need installed before starting are: - Node.js (v16 or higher) - An MCP-capable AI client (Cursor IDE, Claude Desktop, etc.)

npm install @modelcontextprotocol/sdk zod

Set your API key for Agile Luminary as an environment variable. Replace your_api_key_here with your actual key.

export LUMINARY_API_KEY=your_api_key_here

Start the MCP server using Node.js. This launches the server that listens for MCP connections from your AI client.

node server.js

Configuration and usage notes

MCP configuration expects a local stdio server that runs the Node process and exposes the MCP endpoint through the standard input/output transport. The server uses the LUMINARY_API_KEY environment variable for authentication with the Agile Luminary API.

Example configuration for an MCP client (Cursor IDE) can point to a local stdio server with the following setup.

{
  "mcpServers": {
    "agile_luminaries": {
      "command": "node",
      "args": ["server.js"],
      "env": {
        "LUMINARY_API_KEY": "your_api_key_here"
      }
    }
  }
}

Additional security, troubleshooting, and notes

Security: The server handles authentication with the Agile Luminary API locally. Keep your API key secure and do not expose it in client configurations or logs.

Troubleshooting: If you encounter authentication failures, verify that LUMINARY_API_KEY is set correctly in the environment where you start the server. If the server cannot reach Agile Luminary, check your network connectivity and API endpoint access.

Notes: The server exposes three MCP tools that your AI client can invoke to retrieve data from Agile Luminary: getRelatedDocuments, getCurrentWork, and getPastWork. Ensure your AI prompts are clear, for example, “Show me related documents for authentication” or “What past work relates to the payment system?”.

Available tools

getRelatedDocuments

Fetch related documents by a search string to surface relevant documentation and specifications.

getCurrentWork

Retrieve work currently assigned to the authenticated user to show active tasks and user stories.

getPastWork

Search and retrieve completed work related to user stories based on a search term to provide historical context.