JetBrains integration

JetBrains IDEs with AI Assistant support MCP servers starting from the 2025.1 release. This works in IntelliJ IDEA, WebStorm, PyCharm, GoLand, and all other JetBrains IDEs.

Setup

1. Open MCP settings

Go to Settings → Tools → AI Assistant → Model Context Protocol (MCP).

2. Add Oka

Click Add and enter the following configuration:

{
  "mcpServers": {
    "oka": {
      "command": "npx",
      "args": ["-y", "@oka-core/reason"]
    }
  }
}

3. Authenticate

In the AI Assistant chat, use the login tool:

Use the login tool to authenticate with Oka

This opens your browser to id.oka.so where you sign in or create an account. Your credentials are stored locally and refresh automatically.

4. Verify the connection

The MCP settings panel should show oka with a connected status. You can also type / in the AI chat and check for Oka tools.

Usage

In the AI Assistant chat, prompt the agent:

“Use Oka context to check what we know about the database module”
“Record a decision about the API versioning strategy”
“Search Oka for learnings about error handling patterns”

JetBrains-specific notes

Working directory

The JetBrains MCP settings include an optional Working directory field. Set this to your project root if @oka-core/reason needs to resolve relative paths.

Auto-Configure for external clients

Starting with 2025.2, JetBrains IDEs can also act as MCP servers themselves. The Auto-Configure button in MCP settings writes config to external clients (Claude Code, Cursor, Codex, Windsurf) so they can use JetBrains tools. This is separate from Oka — it means you can use Oka in JetBrains while also letting external agents connect to your IDE.

Tips

Set OKA_WORKSPACE in the env block to scope events to your project.
The MCP settings UI supports environment variables alongside command and args — no need to edit config files directly.
AI Assistant needs to be enabled in your JetBrains IDE for MCP to work.