Skip to main content
Version: v1.6.53-dev1 🚧

Import from Confluence

Overview​

The Import from Confluence ingestion method imports content directly from Atlassian Confluence Cloud. The connector supports individual pages, entire spaces, nested sub-pages, and attachments (embedded images).

note

Currently, only images embedded in pages are ingested.

Import from Confluence

When to use​

Use this connector when you want to:

  • Knowledge bases: When your documents are stored in Confluence
  • Team wikis: For collaborative documentation and team wikis
  • Technical specs: For technical specifications and guides authored in Confluence
  • Product documentation: For product documentation managed in Confluence
  • Confluence Cloud: For connecting to Confluence Cloud instances

Required permissions​

  • read:me: Read user profile information.
  • read:page:confluence: Read Confluence pages.
  • read:space:confluence: Read Confluence spaces.
  • read:attachment:confluence: Read embedded images.
  • read:hierarchical-content:confluence: List subpages.
  • offline_access: Refresh tokens without requiring users to sign in again.

Authentication​

The Confluence connector supports two authentication methods: OAuth authentication and API tokens.

API token authentication​

You can create API tokens and store them in the secured store. If you create a scoped token, it must have the scopes listed in the Required permissions section.

OAuth authentication using h2oGPTe data connectors​

h2oGPTe provides secure OAuth authentication that lets you authorize access to your Confluence content without sharing passwords.

To set up OAuth authentication, follow these steps:

  1. In the connector configuration, click Set up OAuth authentication using h2ogpte data connectors to start the OAuth flow. Setup OAuth
  2. On the Data Connectors page, click Connect to authorize h2oGPTe to access your Confluence content. Setup OAuth
  3. On the Atlassian authorization page, click Accept to grant access. Accept Auth
note

After you complete authentication, you can disconnect the Confluence data connector at any time by clicking Disconnect on the Data Connectors page. Disconnect

Technical details​

The Confluence connector integrates with the Atlassian Cloud REST API v2 using:

  • OAuth flow: Atlassian OAuth 2.0 with PKCE
  • Authorization endpoint: https://auth.atlassian.com/authorize
  • Automatic token refresh: Tokens are automatically renewed when they expire
  • User consent: Users explicitly authorize h2oGPTe to access their Confluence content

Importing content​

Import Steps

After authentication, you can import content from Confluence.

To import content, follow these steps:

  1. Select the Credential ID for the authenticated Confluence data connector.
  2. Enter the Confluence URL (instance, space, or page) to specify what to import.
    note

    The connector accepts multiple URL formats. For details, see URL formats.

  3. To include personal spaces, turn on the personal spaces option.
  4. To limit the import to specific spaces, enter space filter keywords.
  5. Click Browse to navigate and select the content to import. Browse Files
  6. After you finish selecting content, click Add selected to start ingestion. Job in progress
note

For document processing options, see the Shared Document Processing Options section.

URL formats​

The connector accepts several URL formats to specify what content to import.

Page URLs​

https://your-instance.atlassian.net/wiki/spaces/SPACE/pages/123456/Page+Title

Imports the specified page. When browsing, you can choose to include nested sub-pages.

Space URLs​

https://your-instance.atlassian.net/wiki/spaces/SPACE/overview

Imports all accessible pages from the specified space.

Instance URLs​

https://your-instance.atlassian.net/wiki

Lists all accessible spaces for browsing and selection.

Page IDs​

You can also use page IDs directly for programmatic access.

Code examples​

Ingest a single page with subpages​

from h2ogpte import H2OGPTE
from h2ogpte.connectors import ConfluenceCredential

client = H2OGPTE(address="http://localhost:8888", api_key="your_api_key")

# Create credentials
credentials = ConfluenceCredential(
    username="your_email@example.com",
    password="your_api_token"
)

# Create a collection
collection_id = client.create_collection(
    name="My Confluence Content",
    description="Imported from Confluence"
)

# Ingest page by ID
job = client.ingest_from_confluence(
    collection_id=collection_id,
    base_url="https://your-instance.atlassian.net/wiki",
    page_id="5094539270",
    credentials=credentials,
)

Ingest multiple specific pages​

# Ingest multiple pages by their IDs
job = client.ingest_from_confluence(
    collection_id=collection_id,
    base_url="https://your-instance.atlassian.net/wiki",
    page_id=["5094342673", "5094473739"],
    credentials=credentials,
)
Feedback