Deploying Your First MCP Module

Overview

In MindCore, MCP Modules are the core logic units that power agents and workflows. This page guides you through deploying your first MCP Module — whether it's a simple data fetcher or a full analytical engine.


Step 1: Define Your Module Logic

Each MCP Module should have:

  • Name – A clear, descriptive title (e.g., “ETH/USD Price Feed”).

  • Input Schema – What data the module expects (e.g., asset symbol, interval).

  • Output Schema – The expected output (e.g., price, percentage change).

  • Execution Logic – The function that processes input and returns output.

You can write your module in Python, JavaScript, or other supported runtimes.


Step 2: Prepare Metadata

Every MCP Module includes metadata to help users understand and use it effectively:

  • Description

  • Author

  • Version

  • Tags (e.g., finance, web-scraping, social-data)

  • Licensing (optional)


Step 3: Register on the Marketplace

  1. Connect your wallet.

  2. Upload the module code and metadata.

  3. Choose your monetization settings:

    • Public (free access)

    • Pay-per-call (usage-based fee)

    • Subscription (recurring fee)

Once submitted, your module is verified (automatically or via community validation) and listed publicly.


Step 4: Test and Monitor

After deployment:

  • Use the Module Runner to test inputs and see live outputs.

  • View logs, call history, and execution stats.

  • Receive usage rewards as other users and agents invoke your module.


Best Practices

  • Keep logic focused. One module = one job.

  • Validate input and handle edge cases.

  • Add clear documentation to help others use it.

  • Update your module with version control if needed.


What’s Next?

Now that you’ve deployed your first module, you can:

  • Integrate it into an MCP-powered Agent

  • Combine it with other modules in workflows

  • Share or monetize it via the marketplace

Last updated