Preparing your workspace…
Preparing your workspace…
Model Context Protocol (MCP) Integration
Connect Kitesurf to your Great Arrow workspace once, and LM Studio (local models) reaches it through GreatArrow.ai's secure, zero-config Model Context Protocol (MCP) server.
Tokens are encrypted via AES-256-GCM. We never store raw credentials in memory or logs.
OAuth sign-in or API key · On demand, no background sync
§01 · Why
§02 · Setup
Sign up at GreatArrow.ai in 30 seconds. No credit card is required to try out the workspace.
Open Connected apps, choose Kitesurf, then: Two ways to connect: Sign in with Cloudflare — click Connect and approve access. We pick up your account automatically. API token — create a Cloudflare API token with the 'Browser Rendering — Edit' permission, then paste it with your 32-character account ID (Cloudflare dashboard → right sidebar → Account ID).
Open integrationsGo to GreatArrow.ai/install, pick your AI, and follow the ~30-second guided connect. LM Studio (0.3.17 or newer) runs open models on your own computer and speaks MCP. Open the Program tab → Install → Edit mcp.json and merge the block below (LM Studio uses Cursor's mcp.json notation), then toggle the server on. The guided installer at https://www.greatarrow.ai/install/lm-studio supplies your workspace token; the public example keeps the placeholder. Your model keeps running locally — only the memory lookups leave the machine.
Open the installer{
"mcpServers": {
"great-arrow": {
"url": "https://www.greatarrow.ai/api/mcp",
"headers": {
"Authorization": "Bearer gad_PASTE_YOUR_TOKEN",
"X-MCP-Client": "lm-studio",
"X-GAD-Install-Version": "2026.11.03"
}
}
}
}§03 · Questions
Ready when you are
Connect your workflows, search historical memories, and automate complex tasks with a secure shared brain.