Intakes overview

Use intakes to bring work from your existing tools into a Factory. Start with a connected repository or issue tracker, then define which requests belong in the backlog. This page helps you choose a source and plan the rules before configuring the workflow.

Choose a source

For your first GitHub repository, follow the Quickstart. For other sources, use the integration components in a custom automation.

Source	Reference
GitHub	Issues and repository events
Jira	Jira issues
Linear	Linear issues and events
Sentry	Sentry events and actions

Component references describe available events and actions. Connecting a source alone does not define how its events become Factory tasks; configure that behavior in your intake automation.

Define what to import

Choose a repository or project and a narrow set of matching requests. Use filters such as labels or issue state where the source supports them.

Each task needs enough context to act on:

• The requested change and expected result.
• The repository and area in scope.
• A link to the original request.
• A test or check that can confirm completion.

Read work intake and backlog for task readiness criteria.

Assess work before starting it

The Backlog automation can inspect a task and record a confidence score with a reason. Use that result to route work, rather than treating it as a guarantee of success.

Keep requests with missing context in the backlog until someone clarifies them. Define separately when implementation may start. See Backlog automation for filtering, assessment and routing rules.

Try one task first

Import one bounded request and confirm its source and context are attached. Check the assessment result before starting implementation. Once the task follows the expected path, expand the intake rules to more work.

Follow Build your first Factory to run tasks through implementation, verification and review.