# MCP Feature Server

## Overview

Feast can expose the Python Feature Server as an MCP (Model Context Protocol) server using `fastapi_mcp`. When enabled, MCP clients can discover and call Feast tools such as online feature retrieval.

## Installation

```bash
pip install feast[mcp]
```

## Configuration

Add an MCP `feature_server` block to your `feature_store.yaml`:

```yaml
feature_server:
  type: mcp
  enabled: true
  mcp_enabled: true
  mcp_transport: http
  mcp_server_name: "feast-feature-store"
  mcp_server_version: "1.0.0"
```

### mcp\_transport

`mcp_transport` controls how MCP is mounted into the Feature Server:

* `sse`: SSE-based transport. This is the default for backward compatibility.
* `http`: Streamable HTTP transport. This is recommended for improved compatibility with some MCP clients.

If `mcp_transport: http` is configured but your installed `fastapi_mcp` version does not support Streamable HTTP mounting, Feast will fail fast with an error asking you to upgrade `fastapi_mcp` (or reinstall `feast[mcp]`).

## Endpoints

MCP is mounted at:

* `/mcp`

## Connecting an MCP client

Use your MCP client’s “HTTP” configuration and point it to the Feature Server base URL. For example, if your Feature Server runs at `http://localhost:6566`, use:

* `http://localhost:6566/mcp`

## Troubleshooting

* If you see a deprecation warning about `mount()` at runtime, upgrade `fastapi_mcp` and use `mcp_transport: http` or `mcp_transport: sse`.
* If your MCP client has intermittent connectivity issues with `mcp_transport: sse`, switch to `mcp_transport: http`.


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://docs.feast.dev/master/reference/feature-servers/mcp-feature-server.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.