Skip to main content

Issues

Early access preview

This feature is in preview phase and is available in Dagster+ in limited early access. Functionality and APIs may change as we continue development. To get early access to this feature, reach out to your Dagster account team. For more information, see the API lifecycle stages documentation.

Issues let you link together multiple failures with a shared root cause, to help you get from noticing a problem to having a fix in production quickly and with full visibility across your team. Issues are designed for an AI-first dev cycle, together with Dagster+ AI and Dagster skills.

Creating Issues​

Issues are automatically created in the Triage state for you by Dagster+ AI through proactive monitoring. You can also create them directly in several ways:

  • From a failed run — Open a failed run and create an Issue to track the failure. You can set a summary yourself, or let Dagster+ generate one for you.
  • From alerts — Alert notifications for failed runs and degraded assets include a link to create an Issue directly from the notification.
  • From a conversation with Dagster+ AI — During a chat session, create an Issue to track the problem. The AI summary and context are attached automatically. For more information, see Chat with Dagster+ AI.

Resolving Issues​

Issues provide context optimized for coding agents. You can resolve them by:

  • Auto-implement with AI — Dispatch an AI agent directly from an Issue to create a fix as a GitHub pull request. See Auto-implement with AI.

  • Using the dagster-expert skill — Invoke the dagster-expert skill from an agent session in your codebase to analyze the root cause and develop a solution. For example:

    /dagster-expert fetch issue 3 and make a plan for resolving the problem

Coming soon​

  • Issue metadata: Set an assignee, priority level, or tags on Issues
  • Project tracker integrations: Integrate directly into your existing workflows through integrations to GitHub Issues, Linear, Jira, etc.