Skip to main content

Custom Tools with MCP

The Model Context Protocol (MCP) is an open standard that enables AI agents to securely interact with external data and tools. While Ordify provides thousands of pre-built integrations via Connect Apps, MCP allows you to connect your agents to your own internal APIs, databases, and custom services.

When to Use MCP​

  • Internal Systems: Connect agents to private databases or internal company APIs that aren't available as public apps.
  • Custom Tools: Build specific logic or functions in your preferred language and expose them to your agents.
  • Data Privacy: Maintain full control over how your data is accessed and processed by the AI.
  • Advanced Integrations: Use MCP when a standard "Connect App" doesn't provide the specific granularity or security your workflow requires.

Setting Up an MCP Server​

To use MCP in Ordify, you first need an MCP-compliant server. You can build your own using the MCP SDKs or use community-provided servers.

1. Add an MCP Server to Ordify​

  1. Navigate to Automation → MCP Servers.
  2. Click Add Server.
  3. Provide a Name and Description for your server.
  4. Choose the Transport Type
  5. Provide the Server URL. If your server requires authentication, you can configure headers or API keys in the configuration section.

2. Test Connection​

Once configured, click the Test Connection button. Ordify will attempt to connect to your server and verify its health.

Discovering and Using Tools​

Once your server is connected and healthy:

  1. Tool Discovery: Ordify automatically queries your MCP server to discover the available tools and their schemas. You can view these tools in the Tools tab of your server configuration.
  2. Assign to Agents:
    • Open the Agent Builder or edit an existing Agent.
    • Navigate to the Integrations section.
    • Your MCP tools will appear under the MCP Servers category.
    • Select the tools you want the agent to use.