🎉 Introducing Auto-Population for the Compass Catalog

[#1]

We’re excited to announce that Compass now auto-populates an accurate, comprehensive software catalog on day one with zero manual setup for all new users.

What’s new?

Compass now automatically populates the catalog by using the Teamwork Graph (TWG)! When you onboard to Compass, you’ll see components for all the repositories in your connected Bitbucket workspace or GitHub organization. Within each of these auto-created components, Compass will attempt to populate the following key metadata:

• Description: Directly derived from the repo description

• Owner team: Inferred based on owner teams assigned in Confluence and Jira or recent contributors to the repository

• Documentation links: Markdown docs and API specs detected in your repository

• Events and activity: Track builds, deployments, merges, and derived metrics from these events

Why does this matter?

An accurate, complete software catalog is a valuable tool in keeping your team in sync to ship faster and more reliably. With auto-population, you now reach that valuable state instantly, with a comprehensive set of documented software components that are enriched with meaningful context.

How does auto-population work?

When your team connects either the Bitbucket integration or the GitHub integration to Jira, your repository information is ingested into the Teamwork Graph. That repository information – along with data from Jira, Confluence, and the rest of your Atlassian apps – automatically creates corresponding components in Compass and populates key metadata such as Owner Team and Documentation. The catalog also updates itself automatically as your repositories change.

This eliminates the need for manual setup or complex integrations, providing instant, accurate visibility into your connected software components.

Get started

1. Connect your Bitbucket workspace or Github organization to Jira:

   • https://support.atlassian.com/jira-cloud-administration/docs/integrate-bitbucket-with-jira/

   • https://support.atlassian.com/jira-cloud-administration/docs/integrate-with-github/

2. Set up a new Compass site: https://support.atlassian.com/compass/docs/set-up-your-compass-site/

3. Open Compass from your site’s navigation (Switch to → More Atlassian products), choose your role, and head straight into your auto-populated catalog.

This feature is automatically enabled for all teams who are new to Compass and have connected Bitbucket or Github to Jira (Standard version or above).

Managing and refining your catalog

After auto-population, you stay in control. You can review, adjust, and expand what Compass creates:

• View all auto-population history in the Component Log view, which you can find on each component page

• Update component details and ownership when needed: https://support.atlassian.com/compass/docs/create-view-update-and-archive-components/

• As you add or remove repositories from your organization or workspace externally, those changes are automatically reflected in your catalog

• Manually create or remove components when appropriate (after auto-population completes)

• If you disconnect a GitHub organization or Bitbucket workspace from Jira, components that were created from those connections are deleted to keep your catalog accurate

Auto-populated data is currently available for new Compass tenants. If you’re on an existing tenant, you can still build your catalog quickly by importing components from repositories linked to Jira: https://support.atlassian.com/compass/docs/import-components-from-repositories-linked-to-jira/

Learn More

You can read more about this feature in the support documentation. We’d love to hear your feedback or help answer any questions. Let us know what you think!

[Oleksandr Sapozhnikov]

What about GitLab integration?

[Cornelia Jeppsson]

How will this work for for newly detected components that are added - we have auto-PRs created for the compass.yml file whenever a new repo is created in Github (and thus a component in Compass) - will that PR now contain a suggestion of ownerID based on who created the repo? (That would be amazing) or how will that work?