Connect to the MCP Server

Learn how to connect Feathery MCP server to Claude Code, Codex, Cursor, VS Code, and other MCP-compatible AI tools using your API key.

This guide shows how to connect an MCP-compatible AI tool to Feathery.

Before You Begin

You need:

• A Feathery API key.

• An AI tool or editor that supports MCP.

• The Feathery MCP URL for your workspace.

For most workspaces, the MCP URL is:

https://api.feathery.io/api/mcp/

If your workspace uses a regional Feathery API URL, use that same API domain and add /api/mcp/.

Find Your API Key

Use a Feathery environment API key. This is different from the SDK key used to embed forms.

Keep the API key private. Anyone with the key may be able to read or update data in the associated Feathery environment.

Add Feathery To Your MCP Client

Most MCP clients ask for:

Field	Value
Name	feathery
URL	https://api.feathery.io/api/mcp/
Authorization header	Token <your Feathery API key>

If your client accepts a full header value, enter:

Authorization: Token <your Feathery API key>

Claude Code

Run:

claude mcp add --transport http feathery https://api.feathery.io/api/mcp/ \
  --header "Authorization: Token $FEATHERY_API_KEY"

Then run /mcp in Claude Code and confirm that Feathery appears in the connected servers list.

Codex

Add Feathery to your Codex MCP configuration in ~/.codex/config.toml:

[mcp_servers.feathery]
url = "https://api.feathery.io/api/mcp/"

[mcp_servers.feathery.http_headers]
Authorization = "Token <your Feathery API key>"

Restart Codex after saving the file. Then ask Codex to list your Feathery forms to confirm the connection is working.

Cursor

Open Cursor's MCP settings and add a new server:

{
  "mcpServers": {
    "feathery": {
      "url": "https://api.feathery.io/api/mcp/",
      "headers": {
        "Authorization": "Token ${env:FEATHERY_API_KEY}"
      }
    }
  }
}

Restart or reload Cursor after saving the configuration.

VS Code

Create or update .vscode/mcp.json:

{
  "inputs": [
    {
      "type": "promptString",
      "id": "feathery-api-key",
      "description": "Feathery API key",
      "password": true
    }
  ],
  "servers": {
    "feathery": {
      "type": "http",
      "url": "https://api.feathery.io/api/mcp/",
      "headers": {
        "Authorization": "Token ${input:feathery-api-key}"
      }
    }
  }
}

VS Code will prompt for your Feathery API key when it starts the server.

Test The Connection

After connecting, ask your assistant:

List my Feathery forms.

If the connection is working, the assistant should return forms from the Feathery environment associated with your API key.