# Configure a Syncfusion MCP server in JetBrains IDEs

Source: https://ai.syncfusion.com/mcp/jetbrains.md
Canonical: https://ai.syncfusion.com/mcp/jetbrains.md
Applies to: IntelliJ IDEA, Rider, WebStorm, PyCharm and other JetBrains IDEs with AI Assistant
Transport: stdio
Credential: MCP API key from https://www.syncfusion.com/account/api-key — not a product license key
Source-confidence: Syncfusion-documented. Syncfusion's MCP documentation lists JetBrains as a supported client, configured through the AI Assistant chat interface rather than a file, and notes the Windows-specific `npx.cmd` command form.

> Before you configure this, read https://ai.syncfusion.com/licensing.md.

## JetBrains differs from the other clients in two ways

1. **Configuration is done through the UI, not a file you write.** Open **Settings → Tools → AI
   Assistant → Model Context Protocol (MCP)** and add the server there. There is no project-level
   `mcp.json` an agent can write on your behalf.
2. **On Windows the command is `npx.cmd`, not `npx`.** This is called out explicitly in Syncfusion's
   documentation and is a common cause of a server that silently fails to start.

## Step 1 — put the key in a file outside the repository

```bash
mkdir -p ~/.syncfusion
printf '%s' 'PASTE_YOUR_API_KEY_HERE' > ~/.syncfusion/api-key.txt
```

On Windows, use a path such as `C:\Users\<you>\.syncfusion\api-key.txt`.

## Step 2 — add the server in AI Assistant

Open **Settings → Tools → AI Assistant → Model Context Protocol (MCP)**, choose to add a server, and
enter:

| Field | Value (macOS / Linux) | Value (Windows) |
| --- | --- | --- |
| Name | `sf-react-mcp` | `sf-react-mcp` |
| Command | `npx` | `npx.cmd` |
| Arguments | `-y @syncfusion/react-mcp@latest` | `-y @syncfusion/react-mcp@latest` |
| Environment variable | `Syncfusion_API_Key_Path=/absolute/path/to/syncfusion-api-key.txt` | `Syncfusion_API_Key_Path=C:\Users\<you>\.syncfusion\api-key.txt` |

Replace `react` with your platform. Server table: https://ai.syncfusion.com/mcp/llms.txt

If the dialog accepts a JSON snippet instead of individual fields, use:

```json
{
  "mcpServers": {
    "sf-react-mcp": {
      "command": "npx",
      "args": ["-y", "@syncfusion/react-mcp@latest"],
      "env": {
        "Syncfusion_API_Key_Path": "/absolute/path/to/syncfusion-api-key.txt"
      }
    }
  }
}
```

## For .NET platforms in Rider

Blazor, ASP.NET Core, ASP.NET MVC, .NET MAUI, WPF, WinForms, WinUI, UWP and Document SDK use NuGet-published servers:

| Field | Value |
| --- | --- |
| Command | `dnx` |
| Arguments | `Syncfusion.Blazor.MCP --yes` |
| Environment variable | `Syncfusion_API_Key_Path=/absolute/path/to/syncfusion-api-key.txt` |

`dnx` requires .NET 10. On .NET 8 or 9 the documented fallback is
`dotnet tool run syncfusion-blazor-mcp`.

## Step 3 — verify

Restart the IDE. In the AI Assistant chat, confirm the Syncfusion tools appear in the tool list, then
ask a question that requires retrieval rather than recall and check that the answer cites
documentation.

## Skills in JetBrains — do not assume a directory

JetBrains is a supported MCP client. It is **not** a documented target for the `skills` CLI, and
there is no verified "JetBrains skills directory". The two JetBrains-adjacent agents the installer
does support are Junie (`.junie/skills/`) and Firebender (`.agents/skills/`). If you are using one of
those, install with `--agent junie` or `--agent firebender`. Otherwise, rely on MCP plus the
documentation, and put the Syncfusion rules in your project instruction file — see
https://ai.syncfusion.com/prompt.md

## If it does not work

| Symptom | First thing to check |
| --- | --- |
| Server never starts on Windows | `npx.cmd`, not `npx` |
| Server not listed | Restart the IDE; AI Assistant reads MCP config at startup |
| Queries fail | API key variable casing: `Syncfusion_API_Key_Path` |
| `dnx` not found in Rider | .NET 10 SDK, or the `dotnet tool run` fallback |

Server table: https://ai.syncfusion.com/mcp/llms.txt
Platform routing: https://ai.syncfusion.com/llms.txt