Integration · SharePoint
Connect SharePoint to meinGPT
Search current SharePoint content directly or index selected libraries as a centrally managed Cloud Vault knowledge source.
SharePoint can be used in meinGPT in two distinct ways. The direct connector reads current sites, pages, folders, and files in the signed-in Microsoft user's context. A SharePoint source in Cloud Vault synchronises selected libraries into a centrally managed search index for larger collections and shared assistants. Both paths are read-only but differ in freshness, permission model, scope, and operation.
With the direct connector, a person authorises SharePoint through Microsoft OAuth and meinGPT calls Microsoft Graph only for the current task. With Cloud Vault, an administrator selects a site, library, or folder and starts synchronisation. The index updates every 15 minutes and uses meinGPT access rules; individual SharePoint permissions are not copied automatically.
Current content in the user's context
meinGPT queries SharePoint through Microsoft Graph for the current task and remains bound to the connected person's Microsoft permissions.
- Best for
- Targeted research in current sites, pages, and document libraries
- Freshness
- Live for each request
- Permissions
- Personal SharePoint permissions apply automatically; an assistant can additionally be limited to selected sites
Shared search across larger collections
Selected SharePoint libraries are synchronised, processed, and indexed for semantic search and shared assistants.
- Best for
- Large document collections, combined sources, and shared knowledge spaces
- Freshness
- Pre-built index synchronised every 15 minutes
- Permissions
- An administrator controls access through meinGPT; individual SharePoint rights are not inherited automatically
Source, permissions, and data residency are documented separately for each connector.
Typical workflows
- Check the current project state
- Make policies searchable for a team
- Combine knowledge from several sources
Compare connection paths
The same system can be connected differently depending on the task. Freshness, permissions, and scope determine the right path.
Direct connector
Current content in the user's context
meinGPT queries SharePoint through Microsoft Graph for the current task and remains bound to the connected person's Microsoft permissions.
- Best for
- Targeted research in current sites, pages, and document libraries
- Access
- Find sites and pages, list files and folders, search content, and load selected files into chat
- Freshness
- Live for each request
- Permissions
- Personal SharePoint permissions apply automatically; an assistant can additionally be limited to selected sites
- Scope
- Whole SharePoint sites; a guaranteed restriction to a subfolder is not available on this path
- Limitation
- The direct connector is read-only and cannot create or modify files, pages, or SharePoint columns.
Setup
- 1.Enable SharePoint in a chat or assistant
- 2.Connect the Microsoft account and optionally define allowed sites
Cloud Vault
Shared search across larger collections
Selected SharePoint libraries are synchronised, processed, and indexed for semantic search and shared assistants.
- Best for
- Large document collections, combined sources, and shared knowledge spaces
- Access
- Search returns relevant passages and source files; selected full text can be loaded when needed
- Freshness
- Pre-built index synchronised every 15 minutes
- Permissions
- An administrator controls access through meinGPT; individual SharePoint rights are not inherited automatically
- Scope
- A site, document library, or selected folder including subfolders
- Limitation
- This is not live access. Changes appear after synchronisation and SharePoint rights must be modelled deliberately in meinGPT.
Setup
- 1.Create a Cloud Vault source in the administration
- 2.Choose SharePoint, define scope, start sync, and review meinGPT access
What the integration makes possible
01
Check the current project state
Use the direct connector to find the latest files in an approved site and load only the relevant documents into the current task.
02
Make policies searchable for a team
Index an approved policy library in Cloud Vault and expose it to a defined group through a shared assistant.
03
Combine knowledge from several sources
Bring selected SharePoint libraries and other cloud sources into one bounded search space with source references.
Clarify before starting
- 01Direct research or a shared knowledge index has been chosen deliberately.
- 02Sites, libraries, or folders in scope are documented.
- 03Owners for OAuth, sync, access, and deletion are assigned.
Review the prompt and result openly
Check the approved SharePoint project site for Project ACME. List the last three documented decisions, open actions, and the source file with its modification date.
Result
A table lists three decisions or actions, their status, source file, and modification date. Results use only the selected SharePoint scope.
Does this connection fit your use case?
We clarify the source, permissions, and connection required for the concrete use case. If the existing integration fits, you can move straight to setup afterwards.
Security and operations
Use the direct connector where personal SharePoint permissions must remain the access model. For Cloud Vault, define source scope and meinGPT groups before the first sync because SharePoint permissions are not automatically reproduced in the index.
Known limitations
- 01The direct connector is read-only and scopes to sites rather than guaranteed subfolders.
- 02Cloud Vault does not automatically inherit individual SharePoint permissions.
- 03The Cloud Vault index reflects the latest completed synchronisation rather than every immediate change.
Frequently asked questions
Use the direct connector for current, targeted research with personal SharePoint rights. Use Cloud Vault for larger collections, combined sources, or shared knowledge spaces.