Model Context Protocol

The Model Context Protocol (MCP) is an open standard that allows AI assistants and developer tools, such as Claude, Cursor, and other LLM-powered clients, to interact directly with external services and APIs in a structured, context-aware way. Instead of an AI assistant simply reading documentation, MCP enables it to actively call your API, understand its capabilities, and execute operations on behalf of the user.

Theneo's MCP Server feature lets you generate a fully configured MCP server directly from your existing API specification. Once generated, your API becomes immediately consumable by any MCP-compatible AI client - without requiring custom integration work on your end.

How it works, in three steps:

1. You configure and generate an MCP server from your API spec in Theneo.
2. Theneo produces a ready-to-run server that exposes your selected API operations as structured tools.
3. An MCP-compatible AI client connects to your server and can discover, understand, and call your API operations in real time.

Accessing MCP Settings

Navigate to the MCP configuration page directly at:

https://editor.theneo.io/mcp

Configuration

Base URL Override

The Base URL Override field allows you to specify a custom API base URL for your MCP server (e.g., https://api.example.com).

Transport Method

Use the Transport Method dropdown to select how your MCP server will communicate. The following options are available:

• STDIO - Standard input/output. Best suited for local development environments and CLI-based integrations.
• HTTP - Standard HTTP transport. Recommended for server-to-server communication and web-based deployments.
• Server-Sent Events (SSE) - Enables real-time, one-directional streaming from server to client. Ideal for use cases requiring live updates.

Select the transport method that best fits your deployment environment and infrastructure requirements.

API Operations

The API Operations section displays all available endpoints from your API specification. Each operation is listed with its HTTP method, path, and a brief description.

• Use the checkboxes to select or deselect individual operations you want to include in the generated MCP server.
• Click Deselect All to quickly clear all selections, or reselect operations individually as needed.

Only the operations you select will be exposed through the generated MCP server.

Authentication & Security

The Authentication & Security section allows you to manage access controls and protect your data by configuring the authentication settings required for your MCP server. These controls ensure that only authorized users can connect and interact with your server safely.

Security schemes are automatically detected from your OpenAPI specification. For each scheme, you can define the Environment Variable Name that your MCP server will use to resolve the corresponding credential at runtime.

Advanced Settings

The Advanced section provides additional configuration options to fine-tune your MCP server's behavior.

Request Timeout (ms)

Defines the maximum time (in milliseconds) the MCP server will wait for a response before timing out. The default value is 30,000 ms (30 seconds). Adjust this value based on the expected response times of your API.

Enable Pagination Hints

Check this option to add pagination support for list-based operations. When enabled, the MCP server will include hints that help AI clients understand how to paginate through results effectively.

Generating the MCP Server

Once you have completed your configuration click the Generate MCP server button located in the bottom-right corner of the page.

Theneo will process your configuration and generate a ready-to-use MCP server based on your API specification and the options you defined.