> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://developers.alephant.io/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://developers.alephant.io/_mcp/server.

# Confluence

> Configure Confluence Cloud or Server and Data Center indexing with explicit scope and permission boundaries.

## What it indexes

The Confluence connector indexes pages, page comments, and supported attachments from Confluence Cloud or Confluence Server and Data Center. Page and attachment hierarchy is retained for navigation and retrieval.

Choose exactly one scope in the connector form:

| Scope      | Indexing effect                                                                                                                          |
| ---------- | ---------------------------------------------------------------------------------------------------------------------------------------- |
| Everything | Indexes every page the credential can read.                                                                                              |
| Space      | Indexes pages in one Confluence space key.                                                                                               |
| Page       | Indexes one page ID, optionally including all descendants.                                                                               |
| CQL Query  | Indexes pages returned by a page-only CQL query; OpenCore adds its own update-time filters and still processes comments and attachments. |

CQL must return only `page` objects. Do not add a `lastmodified` filter because the connector owns update-window filtering. Unsupported or oversized attachments, and images when image processing is disabled, can be skipped.

## Prerequisites

* An OpenCore Admin must know whether the source is Confluence Cloud or Server and Data Center and must have the correct wiki base URL.
* The credential must be able to read every selected space, page, comment, attachment, user, group, and restriction needed by the chosen access mode.
* Confluence Cloud OAuth requires `OAUTH_CONFLUENCE_CLOUD_CLIENT_ID`, `OAUTH_CONFLUENCE_CLOUD_CLIENT_SECRET`, a correct `WEB_DOMAIN`, and the deployment's OAuth state store.
* OAuth is implemented for Confluence Cloud only. Server and Data Center use a manually created personal access token credential.
* For source-permission sync on Confluence Data Center 9.1 or later, use an account that can read REST space permissions; current validation treats the required administrator access as mandatory.

## Credentials

| Method                                      | OpenCore credential values                                                                                     | Notes                                                                                                                                           |
| ------------------------------------------- | -------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------- |
| Confluence Cloud OAuth                      | No token is pasted. After consent, select the accessible Confluence site during finalization.                  | OpenCore stores access/refresh tokens, the selected cloud ID, site name, and wiki URL. OAuth renewal requires the same configured OAuth client. |
| Cloud API token                             | `confluence_username: <CONFLUENCE_USERNAME>` and `confluence_access_token: <CONFLUENCE_API_TOKEN>`             | Cloud manual authentication uses the username and token together.                                                                               |
| Server or Data Center personal access token | `confluence_username: <CONFLUENCE_USERNAME>` and `confluence_access_token: <CONFLUENCE_PERSONAL_ACCESS_TOKEN>` | The Admin form collects both fields; the runtime uses token authentication for non-Cloud instances.                                             |

Enable **Using scoped token** only for a scoped Confluence token. That mode adjusts the API URL and authentication path and uses the compatible spaces endpoint.

## Configure in OpenCore

1. In the Admin panel, open **Connectors**, choose **Confluence**, and create or select the credential.
2. Set **Is Cloud** correctly. OAuth credentials lock this value to Cloud.
3. Enter the wiki base URL, such as `https://<TENANT>.atlassian.net/wiki` for Cloud.
4. Enable **Using scoped token** only when the manual token is scoped.
5. Select Everything, Space, Page, or CQL Query and complete only that scope's fields.
6. Choose the document access mode your deployment supports, set refresh and pruning options, and create the connector.
7. Run the initial index attempt and review skipped attachments or permission failures.

## Permissions

Confluence source-permission sync is a conditional deployment capability. When **Auto Sync Permissions** is available and selected, OpenCore can resolve space permissions, page and ancestor restrictions, users, and groups. Cloud OAuth requests read scopes for content, permissions, users, groups, attachments, search, and offline access.

When source-permission sync is unavailable or not selected, Confluence restrictions are not enforced automatically in OpenCore search. Use **Private** with explicitly assigned OpenCore user groups for restricted knowledge, or **Public** only when all OpenCore accounts may see the indexed pages and attachments.

## Verify

1. Confirm credential probing finds at least one accessible space.
2. Run an index attempt and verify pages from the selected scope appear.
3. Check that a page comment and a supported attachment are searchable when present.
4. Confirm a page outside the selected space, page subtree, or CQL result is absent.
5. If Auto Sync Permissions is selected, test both an allowed and a restricted Confluence user.

## Troubleshooting

| Symptom                                             | Check                                                                                                                    |
| --------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------ |
| No spaces are found                                 | Verify the wiki base URL, Cloud toggle, token type, and credential access.                                               |
| OAuth cannot start or refresh                       | Verify both Confluence OAuth client variables, the callback domain, and the OAuth state store.                           |
| A CQL connector fails or returns unexpected content | Require `type=page`, remove `lastmodified`, and validate the query in the same Confluence instance.                      |
| Attachments are missing                             | Check source permission, file support and size, and whether image processing is enabled for image attachments.           |
| Permission sync fails on Data Center                | Use an account with administrator access to the REST space-permissions endpoint, especially on Data Center 9.1 or later. |

## Related pages

* [Connectors & Indexing](/opencore/knowledge/connectors)
* [Index Settings](/opencore/knowledge/index-settings)