What AI Worker does
AI Worker connects to your GitHub or GitLab projects and runs supported coding agents in your environment. Teams drive work with issue and pull/merge request labels instead of opening a terminal on the worker host for day-to-day tasks.
How label-driven implementation and review work
After you connect a forge and enable repositories, automation is triggered from labels in GitHub or GitLab:
- Add the ai label on an open issue to queue implementation. The agent works from a fresh checkout of the repository’s target branch and, by default, opens a pull or merge request with the result.
- Add the ai-review label on an open pull request (GitHub) or merge request (GitLab) to queue an AI review comment.
- Status labels such as ai-working, ai-pr, ai-failed, ai-reviewing, and ai-reviewed are managed by the worker as jobs progress. You normally only add ai, ai with no-change for question-only answers, or ai-review.
Supported platforms
AI Worker integrates with the platforms supported by the project:
- GitHub
- GitLab.com
- GitLab Self-Managed
Self-hosted deployment and infrastructure ownership
You install and run AI Worker on hardware you control—such as a server, VPS, or private cloud. The software is designed so repositories, forge credentials, and agent runs stay in your deployment. This describes the self-hosted deployment model; it is not a guarantee that repository content or related data never leave your infrastructure in every configuration, especially when agents call external AI providers you configure.
Customers provide and manage their own supported AI provider credentials (for example Cursor, Anthropic/Claude Code, and other agents documented by the project). Those credentials are required for agents to run.
Source-available software (Elastic-2.0)
AI Worker is licensed under the Elastic License 2.0. You can use, modify, fork, and self-host it with the published image and documentation. You cannot offer AI Worker (or a substantial fork) to others as a hosted or managed SaaS. Source availability does not include paid professional services unless you arrange them separately.
Source repository on GitHub · AI Worker Documentation
Professional services available
For organizations that want help deploying or operating a self-hosted setup—especially those using GitLab Self-Managed—paid professional services are available on request. These services are optional and are not included automatically with the software:
- Installation and deployment
- Security hardening
- GitLab Self-Managed integration
- AI workflow and prompt customization
- Optional maintenance and technical support
Engagements are scoped individually. No fixed enterprise price list is published; contact us to discuss your environment and goals.
Important notes
- You supply and manage credentials for the AI providers you choose to enable.
- AI Worker does not guarantee code quality, vulnerability-free output, or complete data isolation.
- Self-hosting keeps the worker under your operational control; agent calls to external AI providers you configure follow those providers’ data handling practices.
- Professional services are scoped per engagement and are separate from the Elastic License 2.0 software license.