How-To Guides
Create a Downstream Connector

How to Create a Downstream Connector for Your MCP Server

This guide walks you through creating a downstream connector in Findexar for your MCP server.

Prerequisites

  • A Findexar organization (see Setup Your Org)
  • An MCP server (Zapier, Make, n8n, OpenAI App SDK, or custom)
  • MCP server endpoint URL and authentication credentials

Understanding Downstream Connectors

Downstream connectors are your backend MCP servers that do the actual work. They can be:

  • Zapier MCP Server - Connect to Zapier automations
  • Make MCP Server - Connect to Make.com workflows
  • n8n MCP Server - Connect to n8n automations
  • OpenAI App SDK MCP - Custom MCP servers built with OpenAI's SDK
  • Custom MCP Servers - Any MCP-compatible service

For more details, see Understanding Connectors.

Step 1: Access Connector Settings

  1. Log in to your Findexar organization
  2. Navigate to your Project and Deployment
  3. Click on Connectors or Downstream Connectors
  4. Click Create Connector or Add Downstream Connector

Step 2: Select Connector Type

Choose the type of downstream connector you want to create:

Zapier MCP Server

  1. Select Zapier as the connector type
  2. Enter your Zapier MCP Server URL
  3. Provide API Key or authentication credentials
  4. Configure connection settings

Make MCP Server

  1. Select Make as the connector type
  2. Enter your Make MCP Server URL
  3. Provide API Key or authentication credentials
  4. Configure connection settings

n8n MCP Server

  1. Select n8n as the connector type
  2. Enter your n8n MCP Server URL
  3. Provide API Key or authentication credentials
  4. Configure connection settings

OpenAI App SDK MCP

  1. Select OpenAI App SDK as the connector type
  2. Enter your MCP Server URL
  3. Provide API Key or authentication credentials
  4. Configure connection settings

Custom MCP Server

  1. Select Custom as the connector type
  2. Enter your MCP Server URL
  3. Provide API Key or authentication credentials
  4. Configure connection settings
  5. Test the connection to verify it works

Step 3: Configure Connection

Configure the connection settings:

  1. Connector Name - Give your connector a descriptive name
  2. Server URL - The URL of your MCP server
  3. Authentication - API key, OAuth, or other authentication method
  4. Connection Timeout - Set timeout for connection attempts
  5. Retry Settings - Configure retry logic for failed requests

Step 4: Test Connection

Test the connection to ensure it works:

  1. Click Test Connection
  2. Verify that the connection is successful
  3. Check that tools are discovered correctly
  4. Review any warnings or errors

Step 5: Save and Verify

  1. Save the connector configuration
  2. Verify that the connector appears in your connector list
  3. Check that tools are available and accessible
  4. Test a tool call to ensure it works correctly

Step 6: Configure Tools (Optional)

After creating the connector, you can configure individual tools:

  1. Navigate to the connector details
  2. Review the Available Tools list
  3. Configure tool settings if needed
  4. Set default parameters or constraints

Troubleshooting

Connection Issues

  • Connection timeout - Check your MCP server URL and network connectivity
  • Authentication failed - Verify your API key or credentials
  • Tools not discovered - Ensure your MCP server is properly configured

Tool Issues

  • Tools not available - Check that your MCP server exposes tools correctly
  • Tool calls failing - Verify tool parameters and server response format
  • Performance issues - Check server response times and optimize if needed

Next Steps

Now that your downstream connector is created:

  1. Create an Upstream MCP Connector - Create the MCP endpoint for ChatGPT
  2. Map Tools - Select which tools to expose in your upstream connector
  3. Set Credit Prices - Configure pricing for each tool

Related Concepts

Setup Your Org on FindexarCreate an Upstream MCP Connector