In short. The SharePoint connector is shipped: it reads, through Microsoft Graph and read-only, the site pages and document libraries of your Microsoft 365, via an application registered in Microsoft Entra. Each document is attached to the knowledge it describes, and critical knowledge without a document becomes visible.
SharePoint is a library, not a diagnosis. It classifies, versions, shares and secures documents at scale, and it does so well. But a library does not tell you which book you are about to lose. An organisation that has filed everything in SharePoint can be exactly as exposed as one that has filed nothing — it simply knows it less.
The document is not the know-how
A forty-page operating procedure may describe an operation that three people can run, or only one. The file is identical in both cases; the risk is not. That is why a knowledge management approach never starts from documents: it starts from knowledge items, from the people who hold them, and from the processes that depend on them. Documents come afterwards, as evidence.
What an organisation running Microsoft 365 does not see
- How many people actually hold each piece of know-how, and at what level of mastery.
- Which scheduled departures — retirement, transfer, end of contract — will tip a piece of know-how into the red zone.
- Which critical knowledge items are described by no document at all, anywhere.
- Who must learn what, by when, and from whom.
Connecting SharePoint: what KnowledgeCapital reads, and how
The SharePoint connector is shipped. It reads your Microsoft 365 through Microsoft Graph, with an application you register in your tenant in twenty minutes, and read-only: no user account, no personal password, no writing.
- SharePoint site pages, with the text of each of their web parts.
- Every file of the document libraries, sub-folders included: Word, PDF, PowerPoint, Excel, text. A file is re-read only when it changed.
- Documents are attached to measured knowledge, searched with their original link, and counted in the red zone: critical knowledge without any document becomes visible.
Two other paths remain useful: pages reachable by an address are also read by the “Website” connector, and a Power Automate flow can push documents to “Your system” through an API key. The complete step by step, with the Microsoft Entra permissions to grant, is in Connecting your systems.
Read next: Mapping critical knowledge: the method.
Try it free for fourteen days — no credit card.
See also: Mapping critical knowledge: the method · Confluence and knowledge management: measuring the risk · Wiki, document management or knowledge management? · SharePoint knowledge base connector: setup and limits · DMS and knowledge management: Alfresco, Nuxeo, CMIS, WebDAV · Intranet knowledge management: map knowledge, no migration
Frequently asked questions
Which permissions must be granted in Microsoft Entra?
An application registered in your tenant, with the Sites.Read.All application permission and admin consent. No user account and no personal password are used.
Which SharePoint content is read?
Site pages, with the text of their web parts, and the files of document libraries, sub-folders included: Word, PDF, PowerPoint, Excel and text. A file is re-read only when it has changed.
Does the connector write to SharePoint?
No. It only reads; your documents stay in Microsoft 365.
SharePoint, Microsoft 365, Microsoft Entra, Power Automate and Microsoft Graph are registered trademarks of Microsoft Corporation. KnowledgeCapital is neither affiliated with nor endorsed by Microsoft.