AkôFlow examples

Runnable workflows, infrastructure fixtures, showcase assets, and experiment tooling for AkôFlow.

The AkôFlow engine repository contains the runtime and product documentation. This repository owns executable examples so their dependencies, generated assets, and validation workflows can evolve independently from the engine.

Repository layout

• simulation/, local/, kind/, slurm/, and real/: execution and infrastructure examples.
• machine-learning/, generative-ai/, document-ai/, media/, rendering/, agentic-workflows/, and scientific-ai/: complete showcase workflows.
• connections/, onboarding/, and server-instance/: reusable connection and installation templates.
• experiments/: Python and Node.js tools used to launch, collect, analyze, and render reproducible scheduling experiments.

Each directory documents its prerequisites. Run a showcase fixture from its own directory with ./run.sh. Most API-driven examples expect an AkôFlow daemon at http://localhost:8080 and accept the usual AkôFlow environment variables for endpoint and token overrides.

Showcase workflow documents declare an OCI image and a structured command for each activity. They never invoke docker run: AkôFlow owns image delivery and container startup for the selected runtime. The local run.sh fixture executes the deterministic workload directly, while CI builds and validates the image as a separate concern.

Product documentation is published at akoflow.com.

New runnable examples follow the showcase standard: offline smoke execution, real computation, observable stages, stable artifact contracts, and optional larger profiles.

Continuous integration

The example files live in this repository, while the reusable CI implementation is maintained in UFFeScience/akoflow/.github/workflows/examples-ci.yml. The small workflow in this repository passes the exact pull-request or main commit to that central workflow. Runnable showcases are discovered from their docker/Dockerfile and executable run.sh; adding a showcase does not require editing a duplicated matrix.

After a push to main, CI publishes validated showcase images to GHCR and notifies the AkôFlow documentation workflow. Configure the repository secret AKOFLOW_DOCS_DISPATCH_TOKEN with access to dispatch workflows in UFFeScience/akoflow. If the secret is temporarily unavailable, the daily documentation reconciliation still checks out this repository and validates all referenced example files.