Does gdrive_root need a Google API key?

No. It reads the folders that Google Drive for desktop has already synced on your Mac. There are no API keys, OAuth scopes, or cloud configuration — if Drive is mounted in Finder, the tool can see it.

Can it see multiple Google accounts?

Yes. If you have more than one Google account signed into Drive for desktop, gdrive_root lists each account's My Drive and Shared drives as separate per-account mounts so the AI can tell them apart.

Does this upload my Drive files anywhere?

No. The tool only lists folder roots locally on your Mac. Nothing is sent to a server or stored in the cloud — all processing happens on your device.

List Synced Google Drive Folders on Mac with AI

The gdrive_root tool lets your AI assistant see every Google Drive folder synced on your Mac — your personal My Drive, any Shared drives you belong to, and the per-account mount points when you have more than one Google account signed in. Instead of digging through Finder, you can just ask. A natural prompt looks like: "List the Google Drive folders synced on this Mac and tell me which accounts are mounted." The AI returns the live folder roots from your real File Stream / Drive for desktop setup, so it always reflects what is actually mounted right now.

This is the discovery step that anchors everything else: once the AI knows where your Drive roots live, it can navigate, read, and organize files inside them. Download LMCP to enable it.

Which AI agents work?

LMCP exposes gdrive_root to every major AI client. Desktop apps auto-configure through a local stdio connection, while web AIs connect through the LMCP Cloud Relay connector:

Desktop (local stdio, auto-configured): Claude Desktop, Cursor, VS Code (GitHub Copilot), Windsurf and Zed.
Web (via LMCP Cloud Relay): ChatGPT, Claude.ai, Grok and Perplexity.

The same tool behaves identically across all of them — the only difference is the transport.

Automation

On its own gdrive_root is a lightweight lister, but it shines as the first link in a chain. A typical flow: call gdrive_root to get the mounted roots, then have the AI browse into a Shared drive, read a spreadsheet or document, summarize it, and drop the summary into Notes or an email — all in one conversation. Because the roots are returned as real paths, the AI can hand them to Finder-style file tools without you copy-pasting anything. Multi-account users especially benefit: the AI can disambiguate "the work Drive" from "the personal Drive" by reading the per-account mounts.

Context

The AI is working with your actual Mac, not a generic cloud snapshot. gdrive_root reads the live mount state of Google Drive for desktop, so it reflects exactly which accounts are signed in and which Shared drives are available at this moment. If you mount a new Shared drive or add a second account, the next call sees it immediately. That grounding in your real environment is what makes follow-up requests accurate instead of guesswork.

Productivity

Hunting for the right Drive folder across multiple accounts is a small but constant tax — open Finder, expand the sidebar, figure out which mount is which. Asking the AI collapses that into a single sentence and a one-second answer. For people juggling a personal and a work Google account, or several Shared drives, it removes the "which one was it?" friction every time you start a file task, and it gives the AI the context it needs to act on the right location the first time.

Privacy & GDPR

Everything runs locally on your Mac. gdrive_root only reads the folder roots that Google Drive for desktop has already synced — nothing is uploaded, copied to a server, or stored anywhere outside your machine. There are no API keys and no cloud middleman. LMCP is one of 150+ tools that are GDPR-compliant by architecture, because your data never leaves the device.