Skip to main content

Set up Omni

Connect docs via MCP
Who can do this?

You probably need your Omni Organization Admin to generate the API token—you may not have access yourself.

Atlan uses Omni's REST API to extract metadata. The connector authenticates with a single bearer token that must be authorized to read across the entire organization—models, topics, folders, dashboards, and workbooks. Omni's API tokens currently inherit the permissions of the user that created them, so Atlan recommends creating a dedicated service user for catalog crawls.

  1. In Omni, sign in as an Organization Admin.
  2. From the left nav, click Settings > Users.
  3. Click Invite user and create a dedicated service user (for example, atlan-service@yourcompany.com).
  4. Assign the Organization Admin role to the new user so it can read every model, topic, folder, and document.
  5. Have the service user complete the email-verification flow and sign in once.

Why a dedicated service user? Omni's API tokens inherit user-level permissions, so a personal-account token grants Atlan only what that user can see. A service user with organization-wide read access ensures the crawl returns a complete catalog and keeps your audit log clean.

Generate the API key​

  1. Sign in to Omni as the service user created earlier.
  2. From the left nav, click Settings > API Keys.
  3. Click to generate a new API key and give it a recognizable name (for example, atlan-catalog).
  4. Copy the generated token immediately—Omni shows it only once. The token has the format omni_osk_.... Save it somewhere secure for crawling Omni.

You also need your Omni API base URL—this is the URL you use to log in to Omni with an /api suffix, for example https://your-org.omniapp.co/api. You enter both the base URL and the token when you crawl Omni.

For end-to-end lineage from warehouse tables through Omni topics to Omni documents, Atlan also needs:

  • An Atlan connection to the upstream warehouse that Omni queries (Snowflake, BigQuery, Redshift, Databricks, or any other supported source).
  • The warehouse connection crawled before or alongside the Omni crawl, so the upstream tables exist in Atlan when Omni's topics are processed.
  • The Omni connection ID for each warehouse mapped to that warehouse connection's qualified name in the Atlan source-connection map field when you crawl Omni. Without this map, the connector has no way to know which Atlan connection backs each Omni topic, so no lineage is emitted.

Without the upstream warehouse connection and the source-connection map, Omni metadata still catalogs cleanly but lineage Process edges from warehouse tables to Omni topics are missing.

What's next?​

You're ready to crawl Omni assets.

Tags:
Was this page helpful?