Install for LM Studio

Run open models on your own machine and give them a memory. Add the server in LM Studio's mcp.json — the model stays local, the memory follows you.

1 Get your link
2 Add it
3 Send a message
· ~30 seconds

You can read the LM Studio setup steps without signing in.

Sign in when you are ready to create a private token, or replace the placeholder in the command below with an existing Great Arrow token.

Try one real project in LM Studio

After connecting, paste this into your AI. It asks about your project, saves your answers, then reads the note back. You don't need to connect files first.

Your first project

Help me make progress on a project with Great Arrow. Check workspace_introspect for a saved first project; reuse it if present, and retry if progress is unavailable. If none exists, ask what I am working on and its next milestone, then save only what I tell you with memory_create, tag "first-project". Use memory_get with the saved id and show me what was stored. Recommend one small next action I can take now and how to check it is done, grounded in that note. For a coding project, use the stated milestone and blockers; do not assume access to my code. Ask before creating a task or reminder.

Step 1 · Get your secure link

The install command below is public, but LM Studio needs its own private link before it can access your Great Arrow workspace. Sign in to create one, or paste an existing link code anywhere you see gad_PASTE_YOUR_TOKEN.

Step 2 · Add it to LM Studio

LM Studio runs open models on your own machine and can connect to this server (0.3.17 or newer). The model stays local and gains a memory that follows you to every other AI tool you use. Docs at lmstudio.ai/docs.

In LM Studio open the Program tab (right sidebar) → Install → Edit mcp.json
Merge the JSON below into the mcpServers object (preserve any servers you already have) and save
Toggle great-arrow on in the Integrations list, then ask your model what it remembers

LM Studio mcp.json (mcpServers block, Cursor notation)json

{
  "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": "2026.08.08"
      }
    }
  }
}

Step 3 · Check your AI is ready

Add the connector in LM Studio (Step 2 above) first.

Once added, send a test message from inside LM Studio (not on this page). Great Arrow will automatically listen for that message and confirm it here.

Common questions

Preparing your workspace…

Try one real project in LM Studio

After connecting, paste this into your AI. It asks about your project, saves your answers, then reads the note back. You don't need to connect files first.

Your first project

Help me make progress on a project with Great Arrow. Check workspace_introspect for a saved first project; reuse it if present, and retry if progress is unavailable. If none exists, ask what I am working on and its next milestone, then save only what I tell you with memory_create, tag "first-project". Use memory_get with the saved id and show me what was stored. Recommend one small next action I can take now and how to check it is done, grounded in that note. For a coding project, use the stated milestone and blockers; do not assume access to my code. Ask before creating a task or reminder.

{ "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": "2026.08.08" } } } }