What file types can onedrive_read_file open?

Text and structured formats — .txt, .md, .csv, .json, .xml and .log — plus several code file types such as .py, .js, .ts and .swift. It reads text content, not binary formats like images or PDFs.

Does it work with both OneDrive Personal and Business?

Yes. LMCP detects the OneDrive folders synced on your Mac (Personal and Business), and the tool can also read any file on your local filesystem by path.

Is my file uploaded anywhere?

No. The file is read locally on your Mac by the LMCP server and shared only with the AI client you connected. Nothing is stored on an LMCP server, and there are no API keys required.

Read OneDrive Files with AI on Mac — onedrive_read

The onedrive_read_file tool lets an AI assistant open and read a text file straight from your OneDrive folder (or anywhere on your local filesystem) on macOS. It handles plain text and structured formats: .txt, .md, .csv, .json, .xml, .log, plus several code file types like .py, .js, .ts and .swift. Instead of opening OneDrive, hunting for the file, and copy-pasting its contents into a chat, you just ask. A real prompt looks like: "Read my OneDrive file Notes/Q3-roadmap.md and give me a three-bullet summary." The AI reads the file in place and answers.

This is part of LMCP, a free, native macOS MCP server with 150+ tools that connect AI assistants to the real apps and files on your Mac. Download LMCP to get started.

Which AI agents work?

Almost every major AI client can use onedrive_read_file through LMCP:

Desktop clients — Claude Desktop, Cursor, VS Code (GitHub Copilot), Windsurf and Zed auto-configure over a local stdio connection. No tunnels, no manual JSON beyond the one-time setup.
Web AIs — ChatGPT, Claude.ai (web), Grok and Perplexity connect through the LMCP Cloud Relay connector, which securely bridges the web client to the server running on your Mac.

Once connected, the tool appears automatically in the client's tool list.

Automation

Reading a file is rarely the end goal — it's the first step in a chain. Because LMCP exposes 150+ tools, the AI can compose them in a single request:

Read a CSV from OneDrive, then draft an email summarizing it with the Mail tools.
Read a Markdown spec, then create matching Reminders or Calendar events.
Pull a JSON config, compare it against another file, and write the diff back into a Note.
Read a .log file and have the AI explain the errors it finds.

You describe the outcome; the AI orchestrates the steps.

Context

The power here is that the assistant works with your actual data, not a vague memory or a stale paste. When you reference "the roadmap doc in OneDrive," the AI reads the current version of that exact file. That grounding means summaries, answers and rewrites reflect what's really on your disk right now — across the OneDrive-Personal and OneDrive-Business folders LMCP detects on your Mac.

Productivity

Think of how often a file lives in OneDrive and the answer you need lives in your AI chat. Without LMCP you context-switch: open the app, locate the file, open it, copy, paste, prompt. With onedrive_read_file that collapses to one sentence. For people who keep notes, datasets, configs and logs synced to OneDrive, it removes dozens of small interruptions a day and keeps you in the conversation where the thinking happens.

Privacy & GDPR

Everything runs locally on your Mac. The file is read by the LMCP server on your own machine and passed only to the AI client you chose to connect — nothing is uploaded to or stored on an LMCP server. There are no API keys to hand over and no cloud copy of your documents. This local-first design makes LMCP GDPR-compliant by architecture: your data never leaves your control by default.