Skip to content
FlexiPedia logo

MCP Servers

Unlock the full potential of your AI models with MCP Servers, a cutting-edge feature that streamlines integration with external tools. Whether you're managing e-commerce platforms, handling customer interactions, or automating data processes, MCP Servers make it effortless to con

Flexito OmniAI
Beginnerv4
Read in:

Listen to an audio overview

A short, AI-generated podcast-style summary of this article.

Unlock the full potential of your AI models with MCP Servers, a cutting-edge feature that streamlines integration with external tools. Whether you're managing e-commerce platforms, handling customer interactions, or automating data processes, MCP Servers make it effortless to connect and enhance your workflows.

Discover MCP Servers

MCP, or Model Context Protocol, is an innovative feature within the AI Hub. It makes connecting AI to external tools straightforward, eliminating the need for complex API endpoint configurations.

Locating MCP Servers

To find MCP Servers, navigate to the AI Hub via the left-hand menu.

Advantages of MCP Servers

  • Facilitates seamless integration with platforms like Shopify, Stripe, Intercom, HubSpot, and others.

  • Operates with natural language queries.

  • Allows direct access to external tools within model prompts.

  • Reduces setup time significantly, from hours to mere minutes.

Both beginners and advanced users will find MCP Servers beneficial. Additionally, users can host their own MCP servers for tailored integrations.

The possibilities are limitless: CRM, billing, search, data lookup, and much more.

Integrating with an MCP Server

  • Connect any custom MCP server, including self-hosted ones.

  • Compatible with OpenAI, Anthropic, and similar platforms.

  • Follow these steps: Name > URL > Optional Authentication > Select Tools > Save > Use in Flows.

Illustration: Connecting Shopify

Initial Step

  • Go to OpenAI Playground → Tools → Add MCP Server.

Here, you can integrate tools such as Shopify, Stripe, Intercom, HubSpot, and others. Copy the generated URL for your integration.

Accessing the OpenAI Dashboard

Input your store URL to generate the MCP server URL. Review any necessary authentication and explore the list of available tools.

Second Step

Adding a Server

To incorporate a server, simply input its name, URL, and any necessary authentication details.

Utilising MCP Servers in Flow Builder

  • Access Flow Builder. → Add an Action block. → Under Integrations, select OpenAI.

  • Opt for "Create Model Response" → Update model to: GPT-4.1-Mini

  • (Recommended for optimal compatibility with MCP tools. Avoid using GPT-4o-mini/4o)

  • Choose MCP Servers, then configure tokens and temperature.

Example Data

In the sample response, you'll find a data list. Identify and choose the correct path, highlighted in the example, which leads to the desired results. Extract the data from this path and store it in a variable.

Assistance and Support

For questions or help with setting up MCP servers, contact our support team or engage with the community.