Quick start

lurq is a hosted service — you don't run a database or a sync to use it. You get an API key, then point your agent at the hosted MCP endpoint.

1. Run the guided installer

npx lurqrun install

The installer:

1. prompts for your API key and validates it,
2. detects your installed assistants (Claude Code, Cursor, Windsurf, VS Code / Copilot, Codex),
3. writes a keyed remote MCP entry for each one.

The entry it writes looks like this:

{
  "type": "http",
  "url": "https://api.lurq.run/mcp",
  "headers": { "Authorization": "Bearer <your-key>" }
}

No database credentials ever touch your machine. Restart your agent afterward so it picks up the new MCP server.

2. Non-interactive install

For scripted setups (CI, dotfiles, provisioning), skip the prompts:

npx lurqrun install-skill --agent claude-code --api-key <key>

Use --agent all to configure every detected assistant at once.

3. Verify it's connected

Ask your agent something dependency-shaped, for example:

"What's the best current form library for React? Use lurq."

You should see it call the recommend tool and return a short, scored list with a dataAsOf timestamp. If nothing happens, confirm the agent was restarted and that the MCP server shows up in its tool list.

Next steps

• Learn what each tool does in the MCP tools reference.
• Prefer the terminal? See the CLI reference.
• Curious how scores are derived? Read How it works.