Give AI Access to OneDrive on Mac (Claude, Cursor, VS Code) — No Graph API
The simplest way to let your AI assistant read and write your OneDrive files without the Microsoft Graph API, OAuth tokens, or Azure AD app registration. Works with Claude Desktop, Cursor, VS Code, ChatGPT, and Windsurf. LMCP works directly with the OneDrive folder already synced on your Mac.
Why Other Solutions Require Graph API and OAuth
Connecting an AI assistant to OneDrive the official way means going through Microsoft's Graph API, which requires:
- An Azure AD app registration — you have to register an application in the Azure portal
- OAuth scopes & consent — configure Files.Read / Files.ReadWrite permissions and grant consent
- Tenant-admin approval — many work/school accounts require an admin to approve the app, and most won't for a personal tool
- Token management — access and refresh tokens to store securely and rotate
- Rate limits & throttling — Graph throttles requests, and large files round-trip through Microsoft's cloud
For someone who just wants their AI to work with files they already have, that's a lot of overhead — and often a hard blocker if your tenant admin won't approve apps.
How LMCP Reads OneDrive Differently
LMCP skips the Graph API entirely. Instead of registering an Azure app and managing tokens, it works directly with the OneDrive folder that's already synced to your Mac.
The OneDrive desktop app keeps a local, always-synced copy of your files on disk — typically under ~/OneDrive or ~/Library/CloudStorage. If you're signed into OneDrive on your Mac, those files are already on your machine, and the OneDrive client keeps them in sync with the cloud both ways.
LMCP reads and writes those files directly. No Azure app, no OAuth scopes, no admin approval, no tokens. When the AI writes a file, OneDrive syncs it back to the cloud automatically.
What You Can Access
Through LMCP, your AI assistant can:
- Browse your OneDrive — list root folders and drill into any directory
- Read files — open documents, notes, and text from your synced files
- Write & update files — create new files or overwrite existing ones (synced back to the cloud)
- Search — find files by name or content across your OneDrive
- Move & rename — reorganize files and folders
- Scope access — restrict the AI to a single subfolder so it only sees what you allow
Unlike read-only integrations, LMCP can both read and write — so your AI can draft a file and save it straight into OneDrive, where it syncs to all your devices.
Keep the AI in a Safe Folder
OneDrive often holds your entire working life. If you'd rather not give the AI the run of everything, LMCP has a built-in scope feature. Just ask:
Only give yourself access to my “Work Documents” folder in OneDrive, nothing else.
LMCP restricts itself to that subfolder (and below) for all future sessions — no config file to edit. It's the easiest way to set up a safe AI workspace folder while keeping personal files off-limits.
How to Install
Download LMCP and install it:
- Open the downloaded
.dmgfile from your Downloads folder - Drag Local MCP to your Applications folder
- Open Local MCP from your Applications folder — it appears in your menu bar
Takes about 30 seconds. Your AI clients are configured automatically.
After installation, restart your AI client so it picks up the new MCP tools:
- Claude Desktop — quit completely (Cmd+Q) and reopen
- Cursor — restart the editor
- VS Code — reload the window (Cmd+Shift+P → “Reload Window”)
- ChatGPT / Windsurf — restart the application
Make sure the OneDrive desktop app is installed and signed in, and that your files are synced locally. LMCP detects your OneDrive folder automatically — no API keys, no app to register.
Example Prompts
Once connected, try these with your AI assistant:
- “What's in my OneDrive Projects folder?” — lists files so you can see what you have
- “Find the budget spreadsheet in my OneDrive” — searches by name across your synced files
- “Read the latest proposal in my Work Documents folder and summarize it” — opens and summarizes a file
- “Save these meeting notes as a file in my OneDrive” — writes a new file that syncs to the cloud
- “Organize my OneDrive Downloads folder by moving PDFs into an Invoices folder” — moves and renames files
Because OneDrive syncs both ways, anything your AI writes shows up on your other devices within seconds — no upload step, no copy-paste.
Limitations
Because LMCP works with the locally-synced folder rather than the Graph API, keep these in mind:
- Files must be synced locally — with Files On-Demand, online-only files download when accessed; files you've explicitly kept online-only need to be available locally
- Requires the OneDrive desktop app — the web-only OneDrive doesn't create a local folder LMCP can read
- Sync timing — changes the AI writes appear in the cloud as soon as OneDrive syncs (usually seconds); keep OneDrive running
- Personal and work accounts — LMCP can access whichever OneDrive folders are synced on the Mac, including multiple accounts
- Office file formats — for editing .docx/.xlsx/.pptx content specifically, use LMCP's dedicated Word/Excel/PowerPoint tools alongside the file tools
Compared to Other Approaches
- Microsoft Graph API + Azure app — full programmatic access but needs an app registration, OAuth scopes, and (often) tenant-admin consent. Best for teams building a real OneDrive integration.
- Zapier / automation tools — connect OneDrive to other apps but still rely on Graph tokens and route your files through a third party.
- LMCP (local sync folder) — read and write, zero configuration, runs entirely on your Mac. Best for individuals who want their AI to work with the files they already have.
Real-World Workflows
The value isn't reading one file — it's having your AI work across your files and connect OneDrive to your other tools.
Draft and file in one step
Draft a project kickoff doc based on the notes in my OneDrive “Project X” folder, and save it back there as kickoff.md.
Your AI reads the source material and writes the result straight into OneDrive — no download, no re-upload.
Find and summarize across files
Search my OneDrive for everything related to the Q3 budget, read the relevant files, and give me a one-page summary of where the numbers stand.
Instead of opening files one by one, your AI searches, reads, and synthesizes them into a single answer.
Cross-app meeting prep
Pull the latest contract from my OneDrive, check related emails, and prep a briefing with open questions for tomorrow's meeting.
Because LMCP also reads your email, calendar, and messages, your AI can pull a complete picture across tools — not just files.
LMCP also connects your AI assistant to email, Microsoft Teams, Slack, Calendar, Contacts, and local files. See the full list of guides or learn more at local-mcp.com.