MCP Server Setup Guide
Set up the Pine Labs MCP Server in your AI client. Choose between API Execution (hosted, recommended) and CLI (local) flows.
Pick the setup path that matches how you want to run the Pine Labs MCP Server.
API Execution (Hosted)
Connect Claude Desktop, VS Code, Cursor, and other MCP clients directly to the Pine Labs hosted MCP endpoint. No infrastructure to deploy.
CLI (Local)
Install the pinelabs-mcp CLI to run an MCP server locally. Includes auto-configuration helpers for popular AI clients.
Prerequisites (both paths)
Sign up and enable API access.
Log in and copy your X-Client-Id and X-Client-Secret from the dashboard.
Claude Desktop, Cursor, VS Code (with Copilot in agent mode), Windsurf, OpenCode, GitHub Copilot, or OpenAI Codex.
The hosted flow needs no local runtime. The CLI flow requires Node.js 18 or higher.
Use API Execution for the fastest setup with the fewest moving parts — Pine Labs hosts everything for you. Use the CLI if you want local control, custom endpoints, or to test in air-gapped environments.