# MCP Setup Guide

The Euno AI Assistant can be integrated with your preferred AI coding assistant (Claude Desktop, Cursor, VSCode Copilot, etc.) using the Model Context Protocol (MCP). This allows you to query your data model directly from within your development environment.

## Prerequisites

Before setting up the MCP integration, you'll need:

1. **API Key**: Contact your Euno administrator to obtain an API key for MCP access
2. **Account ID**: Your Euno account identifier (available in your Euno account settings)

## Configuration by Platform

### Claude Desktop (Pro, Max, Teams and Enterprise)

1. Open Claude Desktop Settings
2. Navigate to Connectors
3. Click "Add custom connector" in the bottom of the screen
4. Give this tool a name (i.e "euno"). In the URL, enter: \`<https://api.app.euno.ai/mcp?account\\_id=\\>\<ACCOUNT\_ID>\&api\_key=\<API\_KEY>. Replace \<ACCOUNT\_ID> and \<API\_KEY> with your specific details.
5. Restart Claude. You will need to go through the authentication flow the first time you use the tool.

### Cursor

1. **Access MCP Settings**
   * Open Cursor settings
   * Navigate to Tools & Integrations section
   * Click "New MCP server"
2. **Configure MCP Server** In the `mcp.json` file that opens, add:

   ```json
   {
     "mcpServers": {
       "euno-assistant": {
          "type": "http",
          "url": "https://api.app.euno.ai/mcp",
          "headers": {
              "x-api-key": "<API_KEY>",
              "x-account-id": <ACCOUNT_ID>
          }
        }
     }
   }
   ```

   Make sure to replace \<ACCOUNT\_ID> and \<API\_KEY> with your specific details.
3. **Restart Cursor**

## Available Tools

| Tool                                     | Description                                                                                                          |
| ---------------------------------------- | -------------------------------------------------------------------------------------------------------------------- |
| euno\_instructions                       | Get detailed instructions on how to use the Euno MCP server, including resource types, relationships, and properties |
| fetch\_single\_resource                  | Retrieve properties of a single resource by its URI                                                                  |
| find\_resource\_by\_name                 | Locate resources by name using similarity/semantic search                                                            |
| find\_resources\_for\_topic              | Search resources by topic/keywords across names and descriptions, ranked by relevance                                |
| get\_upstream\_lineage                   | Retrieve upstream lineage (dependencies) for a resource, grouped by relationship type                                |
| resource\_impact\_analysis               | Get downstream impact summary with counts and sample affected resources                                              |
| documentation\_search                    | Keyword search over Euno documentation                                                                               |
| documentation\_get\_full\_document       | Fetch full documentation page by URL                                                                                 |
| documentation\_get\_surrounding\_context | Fetch adjacent doc chunks around a specific chunk ID                                                                 |
| generate\_eql\_query                     | Convert natural language to EQL query                                                                                |
| execute\_eql\_query                      | Run an EQL query to search/filter resources with pagination and sorting                                              |
| execute\_eql\_count                      | Count resources matching an EQL query, optionally grouped by property                                                |

***