OneNote MCP. Search, map, and extract content from your entire digital brain.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Just plug in your AI agents and start using Vinkius.
OneNote MCP Server connects your AI agent directly into all your personal and enterprise Microsoft OneNote notebooks. It gives your client the ability to search, list structures, read raw page content, and append new notes without you ever leaving the chat window.
Stop navigating tabs; just ask your AI client anything about your digital archives.
What your AI agents can do
Get notebook
Fetches detailed properties for a single specified notebook.
Get page content
Pulls the raw, written text from a specific page ID.
List notebooks
Lists all active notebooks in your entire OneNote account.
The agent lists every primary notebook container available in your account using list_notebooks.
You can traverse the full organizational map by listing section groups and then sections within a specific notebook (list_section_groups, list_sections).
The agent lists all individual pages inside any given section using list_pages for structural review.
Use get_page_content to pull the clean, raw text from a single page ID, ignoring proprietary formatting.
The agent executes search_pages to find matching keywords across all notebooks simultaneously, regardless of how deep they are filed.
Retrieve basic properties and configuration details for a specific notebook using get_notebook.
Ask AI about this MCP
Supported MCP Clients
Gemini Waiting for input…
OneNote MCP Server: 7 Tools for Note Retrieval
These seven tools give your agent fine-grained control over reading the structure and fetching specific textual data from every part of your OneNote environment.
019d75e5get notebook
Fetches detailed properties for a single specified notebook.
019d75e5get page content
Pulls the raw, written text from a specific page ID.
019d75e5list notebooks
Lists all active notebooks in your entire OneNote account.
019d75e5list pages
Lists metadata for pages contained within a specific section.
019d75e5list section groups
Maps the main groupings (like 'Q1 2024' or 'Client A') inside a notebook.
019d75e5list sections
Lists all individual sections that act as folders within a parent notebook.
019d75e5search pages
Searches for specific keywords across every available page in all connected notebooks.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with OneNote, then connect any of our 4,700+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 4,700+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
What you can do with this MCP connector
Your AI agent connects straight into all your personal and corporate Microsoft OneNote stuff. You'll get access to every notebook, every section, and every page—all without ever having to click around in a browser tab. This lets you treat the whole mess of digital notes like one giant database. Stop sifting through menus; just ask your agent anything about what you saved.
The server gives your client several specific jobs it can run for you:
Mapping Your Whole Archive
You're going to start by listing every single primary notebook container sitting in your account using list_notebooks. From there, the agent lets you map out the deep organizational structure. You can use list_section_groups to see the major groupings—like 'Q1 Projects' or 'Client Accounts'—that sit inside a main notebook. After that, it pulls through all individual sections within those groups using list_sections.
If you need to know what pages are in a specific folder, you can run list_pages, which shows all the metadata for every page contained within a section.
To get basic details on one specific book of notes, you just use get_notebook to fetch its properties and configuration info. This lets you verify exactly what that notebook is set up to do.
Finding Stuff Fast
The biggest win here is the global search. You don't have to wait for manual indexing; your agent executes search_pages, which finds specific keywords across every single page in all connected notebooks, no matter how deeply they’re filed away. If you know what you’re looking for, it finds it.
When you find a promising-sounding page ID or reference number, you use get_page_content. This tool pulls the clean, raw text straight out of that specific page, completely ignoring all that proprietary formatting junk OneNote loves to throw at ya. You get the actual words you need, nothing more.
How It Works For You
It’s straightforward. Your client sends a command referencing one of these tools and providing the necessary parameters—like listing notebooks or searching for 'Alpha Project.' The server runs that tool call against your OneNote data and sends the resulting structured information back to your chat window. Everything you need is right there, actionable text.
You're not limited to reading; you can use this structure to guide your agent in doing more with the content you find.
How OneNote MCP Works
- 1 First, enable the OneNote MCP integration on your client platform.
- 2 Second, provide an active Microsoft Graph Access Token scoped to OneNote data access.
- 3 Third, send a plain text command in your chat interface (e.g., 'list all notebooks') that triggers the tool.
The bottom line is: you write natural language requests, and the agent translates them into specific API calls against your OneNote data.
Who Is OneNote MCP For?
Project Managers who deal with massive meeting archives; researchers needing to cross-reference thousands of scattered citations; or EAs drowning in client notes. If you spend time jumping between tabs just to find one piece of context, this is for you.
Uses search_pages and list_notebooks to pull obscure citations or related concepts from years of academic notes without manual keyword guessing.
Runs structural queries (list_sections, list_pages) to validate project documentation completeness before a review meeting. Uses get_page_content for immediate status checks.
Uses the agent to quickly pull and append context (like summarizing an email chain) into existing executive notebooks, keeping everything in one place.
What Changes When You Connect
- Bypass manual indexing waits. Use
search_pagesto find keywords across all notebooks instantly, no matter how deeply filed they are. - Get clean text every time. Instead of wrestling with proprietary HTML tags,
get_page_contentextracts the raw written data you need. - Map your organization without clicking through menus. Use
list_notebooksand subsequent listing tools (list_section_groups, etc.) to see the full structural map in text format. - Append contextually generated notes. You can send summaries or quick thoughts directly into an existing section using the agent, keeping your workflow conversational.
- Get a bird's-eye view of your data. By running
list_pagesand related tools, you gain visibility into every single page ID in a target area.
Real-World Use Cases
Finding the right citation for a paper.
A researcher needs to confirm if 'Project Chimera' was mentioned in any notes from 2019. The agent uses search_pages and targets 'Project Chimera'. It finds three matches, one of which is located in a deeply nested notebook structure that the user couldn't find manually. The AI then uses get_page_content to pull out the exact quote.
Creating an updated project timeline.
A PM needs to know what sections cover 'Q4 Milestones'. They first run list_notebooks to confirm the right file. Then, they use list_sections inside that notebook to narrow it down. Finally, they ask the agent to compile all content from those specific sections and append a summary of current status.
Onboarding a new team member.
A manager needs an overview of 'Client X's history'. The agent first runs list_section_groups to identify all client-related containers. It then uses the list IDs to fetch content from the last four sections, providing the new hire with a summarized context dump.
Reviewing old meeting notes for key decisions.
The user only remembers a phrase: 'Use OAuth flow by May.' Instead of searching keywords generally, they ask the agent to search using search_pages and filter results by notebook. This pinpoints the exact page containing the decision, solving the problem instantly.
The Tradeoffs
Trying to dump every piece of data at once.
Asking 'Give me all my notes' will overload the system and only return generic metadata without usable text. You can't read raw content just by listing things.
→
Don't ask for everything at once. First, run list_notebooks to select the container you care about. Then, use list_sections inside that notebook to narrow the scope before running any retrieval tool.
Forgetting where the data is located.
When you know a page exists but don't have the ID, just searching keywords might return too many irrelevant hits, making it hard to pinpoint the source document.
→
If you suspect an area, run list_pages first. This provides structural IDs and titles for that specific section, helping your agent focus its search on a smaller, more manageable set of documents.
Ignoring the structure entirely.
Sending a vague prompt like 'What did we talk about last month?' is too broad. The AI client needs boundaries to find accurate answers and won't guess which notebooks apply.
→
Always scope your request. Start by using list_notebooks to confirm the right project container, then focus subsequent calls (like search_pages) only within that confirmed notebook ID.
When It Fits, When It Doesn't
Use this server if your primary need is reading data from existing notebooks—specifically when you need a structured way to locate content and extract raw text. It's perfect for deep research or documentation retrieval.
Don't use it if: 1) You are trying to create new notes without using the agent's append feature; those edits require different APIs. 2) Your data is in external sources (like SharePoint lists or databases); you need a connector specific to that source instead. 3) You only care about file attachments (like PDFs): this tool handles text content inside OneNote, not files linked outside the page body.
When you need structure traversal and deep reading power, this is your tool.
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by OneNote. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This server provides 7 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.
Available Capabilities
Manually finding context across notebooks shouldn't take 20 clicks.
Right now, if a project note spans three different sections—say, 'Goals,' 'Execution,' and 'Review'—you have to open the notebook. Then you click on Goals, copy the relevant text, switch tabs to Execution, copy those details, then repeat for Review. You end up with a messy document assembled through tedious manual copying.
With this MCP server, your AI agent handles that whole process in one chat exchange. Just ask it: 'Gather all key decisions from Q3 across Goals, Execution, and Review.' It runs the necessary listing tools (`list_sections`, `list_pages`) and pulls clean content using `get_page_content`—it hands you the compiled answer.
OneNote MCP Server: Access raw text with one prompt.
The biggest time sink is dealing with proprietary formatting. When a page uses complex internal tags, just dumping it into plain text messes up the data. You spend time cleaning up what should be clean input.
This server bypasses that mess. By using `get_page_content`, you get the raw written text directly—no extra cleanup needed. It's immediate, reliable context transfer.
Common Questions About OneNote MCP
How do I find a specific keyword in all my notes using search_pages? +
Use search_pages and provide the exact keyword or phrase you need. The tool will scan every page ID across your connected notebooks and report back exactly where it found the match.
Can I list all my client notes using list_notebooks? +
Yes, running list_notebooks gives you a full inventory of all major containers. You can then pass these IDs to subsequent listing tools to map the structure.
What is the difference between list_pages and get_page_content? +
Use list_pages when you just need structural metadata (titles, dates, page ID). Use get_page_content when you actually need the body text written on that specific page.
How do I append a summary to an existing section? +
You tell the agent which Notebook and Section ID to target, provide your summary in the prompt, and the agent handles appending that content directly into the live notes.
How do I use list_sections to traverse deeply nested sections within a notebook? +
list_sections provides a clean, structured view of all folders inside a given container. It requires the parent Notebook ID so it knows which organizational tree you're looking at. This lets your agent build a full map of your document structure before reading any specific content.
When I use get_page_content, how do I handle proprietary Microsoft Graph HTML formats? +
The tool handles the raw data extraction from complex MS Graph HTML structures. When you call it, your agent receives the underlying text and data, stripped of most proprietary tags. This means you don't have to worry about parsing nested formatting codes manually.
What is the primary purpose of get_notebook if I already know the notebook ID? +
get_notebook pulls detailed metadata and configuration properties for a single notebook. Use it when you need to validate permissions, check ownership status, or confirm specific structural details before running large-scale operations on that notebook.
If I use list_pages repeatedly, how should my agent handle rate limits or massive page counts? +
The agent handles pagination automatically when listing many pages. If you're working with thousands of pages, always ask the agent to process results in manageable batches rather than requesting everything at once. This prevents hitting API call limits.
Can the integration delete entire extensive notebooks or important local sections? +
No. The integration exclusively binds heavily to Reading methods (list, search, get) mapped safely alongside minimal Write interactions specifically scoped to Appending fresh notes. Destructive end-points are intrinsically restricted protecting vital long-term data persistently.
Does it parse images contained inside OneNote native pages automatically? +
Currently, fetching Page contents operates exclusively translating returned DOM hierarchies returning raw string-based text bounds flawlessly. Purely visual bitmaps or native handwritten encodings skip the explicit parsing stream avoiding LLM hallucinations fundamentally.
Why does OneNote demand a generic Graph Access Token explicitly here? +
Microsoft explicitly deprecated discrete independent silo APIs unifying all native core tenant structures explicitly over the single Microsoft Graph perimeter gateway. OneNote resources are canonically navigated traversing this exact Graph endpoint hierarchy natively.
Use it with your favorite AI tools
Connect this server to Cursor, Claude, VS Code, and more.
More in this category
YouTube
Search for videos, retrieve detailed statistics, and audit channel performance on YouTube — the world's leading video sharing platform.
Confluence
Empower your AI agent to search, read, and create Confluence wiki pages directly from chat.
Microsoft App Store
Manage your Microsoft Store apps — track submissions, add-ons, and package flights via the Submission API.
You might also like
OpenFEC (Federal Election Commission)
Access real-time federal campaign finance data — search candidates, track committee filings, and analyze election financial totals directly.
Thinkific Alternative
Manage Thinkific groups and webhooks — delete specific student groups and monitor site events directly from your AI agent.
AccuLynx
Business management for roofing contractors — manage leads, jobs, and financial data via AI.