Interact with SharePoint sites, lists, and documents

SharePoint guMCP Server

Search SharePoint sites across the tenant that the user can access

Pagination cursor from a previous response's next_cursor

Free text search to match site names and properties (defaults to all sites)

list_sites

Expected JSON keys: `items` (site objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

Get a SharePoint site's details by site ID or site URL

Preferred: full URL of the site (e.g. 'https://contoso.sharepoint.com/sites/marketing')

get_site

Site object with id, name, displayName, webUrl, and timestamps

list_subsites

Expected JSON keys: `items` (subsite objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

List Microsoft 365 users, for resolving people in list item person fields

Filter to the user with this email address

Filter to users whose display name starts with this text

list_users

Expected JSON keys: `items` (user objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

List all lists in a SharePoint site, or get one list when list_id is given

list_lists

When list_id is given: a single list object. Otherwise expected JSON keys: `items` (list objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

Column definitions, each a columnDefinition object (e.g. {"name": "Priority", "text": {}})

List template, e.g. 'genericList' or 'documentLibrary'

create_list

List the column definitions of a SharePoint list

list_columns

Expected JSON keys: `items` (column definitions with the internal `name` used when reading/writing item fields), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

List items in a SharePoint list, or get one item when item_id is given

OData filter on fields, e.g. "fields/Status eq 'Open'" (best on indexed columns)

list_list_items

When item_id is given: a single item object with its `fields`. Otherwise expected JSON keys: `items` (item objects with `fields`), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

Column values keyed by internal column name, e.g. {"Title": "New item", "Status": "Open"}

create_list_item

Update the field values of an existing SharePoint list item

Column values to set, keyed by internal column name, e.g. {"Status": "Closed"}

update_list_item

delete_list_item

List the document libraries (drives) of a SharePoint site

list_drives

Expected JSON keys: `items` (drive objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

List files and folders in a document library, a folder, or a folder path

ID of a folder to list (defaults to the drive root)

Path of a folder relative to the drive root, e.g. 'Reports/2026'

list_files

Expected JSON keys: `items` (driveItem objects; folders have a `folder` field, files have a `file` field), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

Search for files within one document library

Limit the search to this folder (defaults to the whole drive)

search_files

Expected JSON keys: `items` (matching driveItem objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

Get a file's metadata and download URL by item ID or by its SharePoint URL

Preferred: the SharePoint URL of the file

get_file

The driveItem object; `@microsoft.graph.downloadUrl` is a short-lived direct download link.

Download a file from SharePoint into Gumloop storage

If true, stores the file in workspace storage. Default: false.

download_file

Download result with the stored filename, original name, item ID, and size in bytes

Upload a file from Gumloop storage to a SharePoint document library

Name for the file in SharePoint (defaults to file_name)

If true, reads the file from workspace storage. Default: false.

ID of the destination folder (defaults to the drive root)

upload_file

ID of the parent folder (defaults to the drive root)

create_folder

Copy a file or folder, optionally to another drive or folder

Destination drive ID (defaults to the source drive)

copy_file

Copy is asynchronous. Returns status 'accepted' and a monitor_url that reports copy progress; the copied item appears in the destination shortly after.

move_file

Delete a file or folder (moves it to the recycle bin)

delete_file

Create a sharing link for a file or folder

create_sharing_link

The created permission with the sharing link

list_file_versions

Expected JSON keys: `items` (version objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

restore_file_version

Check out a file to prevent others from editing it

checkout_file

Check in a file, making its latest changes available to others

checkin_file

List the pages of a SharePoint site, or get one page with its content when page_id is given

ID of a single page to fetch with its full canvas content

list_pages

When page_id is given: a single sitePage object with canvasLayout. Otherwise expected JSON keys: `items` (page objects), `next_cursor`. Pass next_cursor back as cursor to fetch the next batch (null when no more).

File name of the page, e.g. 'my-page.aspx'

create_page

The created sitePage object (created in draft; use publish_page to make it visible)

Publish the latest version of a page, making it visible to all users

publish_page

Publish result with the page's publishing state (`published` is true once the page is live)

SharePoint

Installation

Set up the SharePoint MCP server in Gumloop

Create a Gumloop account

Add and authorize the SharePoint server

Then use it in your client

Use SharePoint in an agent

Ship SharePoint agents in minutes