Skip to main content
Every object in EigenPal is a portable artifact: workflows, datasets, evaluators, and agent source are all plain files. That gives you standard developer workflows:
  • Version them in Git and diff exactly what changed between two workflow versions.
  • Review them in pull requests, the same as any code change.
  • Export them between environments with the same commands.
  • Generate them programmatically from code, scripts, or agents.
Nothing is locked inside a UI. You can pull a workflow, change one step, push a new version, run it against a dataset, and read the scores through files and the CLI or API.

​
The artifacts

Every artifact round-trips: what you pull is what you can push back. A single-example dataset export re-imports cleanly, which is useful for moving one failing case between environments.

​
What this enables

  • Files are scriptable. Workflows, datasets, and evaluators can be edited by humans, generated by scripts, or changed by coding agents.
  • Git review works. Diff a workflow change in a PR. Roll back a dataset. Track evaluator history.
  • Environment moves are explicit. Export, inspect, and move artifacts between local, staging, and production with the same commands.

​
The CLI is the surface

The eigenpal CLI covers authoring, running, and inspecting from a terminal. See the CLI reference for the full command tree and the skill install guide for agent-tool setup.