Skip to content

Getting started

Choose how you want to use AntFlow: run workflows on an existing instance, deploy or operate a platform, publish workflows, or contribute to the project.

Use an existing instance

You need the instance's URL, an account accepted by its sign-in provider, and access to a workflow and an available execution environment. Ask your platform administrator for these details. You do not need to install AntFlow or write CWL to run an existing workflow.

For a first look, open the hosted demo. Available projects and execution resources depend on the instance and your access.

  • Start with the User guide.
  • Follow Run a workflow to submit a job, monitor its progress, and retrieve results.
  • For command-line access, follow Access the API to choose the demo or your own instance, configure credentials, and verify access before using the API examples.

Deploy your own instance

The official deployment method is Helm on Kubernetes. You need a cluster, persistent storage, an identity provider, and S3-compatible storage for results. The default deployment includes the Server, UI, a Kubernetes launcher, and PostgreSQL.

Follow Installation with Helm for the requirements and deployment steps. Use the Operator manual to configure access, storage, networking, and execution resources.

Publish your own workflows

Workflow authors describe processing in Common Workflow Language (CWL), prepare the runtime container images, and store the application package in GitLab. AntFlow discovers eligible workflows and exposes them as processes that users can execute.

If you are new to CWL, follow Create your first workflow. For an existing application package, follow Publish a workflow. Your instance's administrator can confirm the available execution environments, repository-access requirements, and project topic.

Contribute to AntFlow

Use the Developer guide to find the contribution instructions and local checks for the Helm chart and documentation.