# VS Code

> Connect GetIntel's MCP server to VS Code's agent mode with a remote HTTP server in mcp.json, so Copilot Chat can read your AI visibility data.

Source: https://getintel.ai/docs/mcp/clients/vs-code/

<Steps>

1. **Create a token** in GetIntel under **Integrations**.

2. **Add GetIntel** to `.vscode/mcp.json` in your workspace:

   ```json title=".vscode/mcp.json"
   {
     "servers": {
       "getintel": {
         "type": "http",
         "url": "https://app.getintel.ai/mcp",
         "headers": { "Authorization": "Bearer YOUR_TOKEN" }
       }
     }
   }
   ```

3. **Start the server** from the **Start** link above the entry, then open Chat in **Agent** mode. GetIntel's tools appear in the tools picker.

</Steps>

<Aside type="caution">
Don't commit a real token. Keep `.vscode/mcp.json` out of version control, or use VS Code's input variables so the token is prompted for instead of stored.
</Aside>

## Or connect without a token

VS Code also supports sign-in. Add the server with no `headers`:

```json title=".vscode/mcp.json"
{
  "servers": {
    "getintel": {
      "type": "http",
      "url": "https://app.getintel.ai/mcp"
    }
  }
}
```

Start the server and VS Code opens GetIntel's sign-in page, then an approval page where you pick a brand and choose whether to allow actions. The access token it gets lasts 2 hours and refreshes itself automatically.