> ## Documentation Index
> Fetch the complete documentation index at: https://docs.letsflow.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Knowledge overview

> Connect knowledge sources and search content people can already access.

Knowledge integrations bring documents from external sources into Flow. People
search the synced copy of files they can already open in the source, after Flow
has stored that grant. When enabled, search can also include connected Jira
issues and web results.

## Capabilities

* Knowledge source bindings (Confluence, SharePoint, Google Drive, ServiceNow KB, and Notion where enabled)
* Scheduled and on-demand sync
* Search across synced content
* Optional Jira and web search
* Source status and sync details in Flow

## Configuration

Knowledge sources are connected per team under **Integrations**. That
team setting is how you pick the site, folder, space, or knowledge base and
watch sync health. It is not the read lock.

| UI step | Required | Purpose |
| - | - | - |
| Provider connect | Yes | OAuth or instance authorization: see provider pages |
| Knowledge source | Per team | Site, drive, space, or KB content to sync |
| Search settings | Optional | Jira and web search options where available |

## Connect a knowledge source

<Steps>
  <Step title="Open Integrations, filter to knowledge integrations, and choose a provider." />

  <Step title="Connect the provider and select the team and content to sync." />

  <Step title="Run an initial sync and confirm that the intended content appears in Flow." />

  <Step title="Enable Jira or web search in product settings if those options are available for your deployment." />
</Steps>

## Scopes and permissions

The provider connection needs read-only access to the content you want to sync. Write access is not required. See the individual provider pages for provider-specific requirements.

For SharePoint, SharePoint Scoped, Google Drive, Dropbox, Box, and
ServiceNow, Flow authorizes reads from the crawled provider access list. A
person can read a file in Flow when the source already grants them access and
Flow has stored that grant. Connecting the source to their team is not required
to read. Connecting it does not grant read. Organization admins cannot bypass
that source access.

When Flow cannot determine who the source allows, it does not grant everyone
access. Some SharePoint pages stay limited to the team that connected the
source. ServiceNow can use an explicit team or organization setting. See
[SharePoint](/integrations/knowledge/sharepoint) and
[ServiceNow KB](/integrations/knowledge/servicenow-kb).

Org-wide **Knowledge** and org-wide search use only that source access.
Team-scoped **Knowledge** and team-scoped search also require a source that
team owns or later connected.

For Confluence Cloud, an administrator chooses at connect time. One choice
mirrors Confluence permissions, including restricted pages, and needs an
Atlassian organization API key. The other choices use a Flow publication
audience: the teams that connected the source, or the organization. Those
publication choices skip restricted pages and do not use the organization API
key. See [Confluence Cloud](/integrations/knowledge/confluence).

For Notion, the connection determines what Flow can crawl. Notion does not give
Flow a reliable list of people who can view each page, so Flow does not copy
Notion Share-menu, inherited team-space, group, guest, public-link, or per-user
permissions. An administrator chooses the Flow publication audience: unpublished,
the teams that connected the source, or the organization. A second team can
connect the same Notion source. Its members read through that publication
audience. They do not receive a Notion people snapshot. Missing or stale
authorization state denies access.

Notion crawl scope follows the root selected in Flow:

* **Page:** the selected page and its structural descendants, including child pages and child databases.
* **Database or data source:** every row plus structural descendants beneath those rows.
* **Workspace (when explicitly configured):** all content that the connection can discover.

To choose a root, search the content shared with the connection, then open a
result to browse what a crawl rooted there covers. A page lists its child pages
and databases. A database lists its data sources, or its row pages when Notion
does not return data sources. A data source lists its row pages. Select any item
at any level. Search always spans the whole connection; there is no search
inside an opened item.

Flow does not expand a selected root through ancestors, siblings, relation properties, mentions, links, or other content merely available to the connection.

## Provider pages

| Provider | Page |
| - | - |
| Confluence Cloud | [Confluence](/integrations/knowledge/confluence) |
| SharePoint Online | [SharePoint](/integrations/knowledge/sharepoint) |
| Google Drive | [Google Drive](/integrations/knowledge/google-drive) |
| ServiceNow Knowledge Base | [ServiceNow KB](/integrations/knowledge/servicenow-kb) |
| Search | [Search](/integrations/knowledge/search) |

## Related topics

* [Product: Knowledge](/setup/knowledge)
* [API Reference](/api-reference)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.