Documentation Development¶
OpenEA Community documentation is maintained in the same repository as the application so product changes and documentation changes can be reviewed together.
Published documentation is available at:
https://docs.openea.dev/
Source layout¶
docs/ Markdown source
mkdocs.yml navigation and site configuration
requirements-docs.txt documentation-only Python dependencies
Do not edit generated HTML under site/. The site/ directory is a local build artifact.
Install MkDocs tooling¶
From an activated development environment:
python -m pip install -r requirements-docs.txt
Preview locally¶
Start MkDocs in the background:
make docs
Default preview URL:
http://127.0.0.1:8000/
If the OpenEA application already uses port 8000:
make docs DOCS_PORT=8001
Check status:
make docs-status
Stop the server:
make docs-stop
Validate before commit¶
Run:
make docs-build
This executes:
mkdocs build --strict
When adding a new documentation page, also add it to mkdocs.yml if it belongs in the published navigation.
Publishing¶
Documentation publishing is automated through GitHub Actions.
edit Markdown / mkdocs.yml
↓
preview and build locally
↓
commit / merge to main
↓
GitHub Actions builds MkDocs
↓
GitHub Pages deploys
↓
https://docs.openea.dev/
The generated site is not committed to the application repository.
Writing conventions¶
- Treat the current OpenEA Community implementation as the source of truth.
- Keep Community documentation independent from OpenEA Enterprise capabilities.
- Link to reference pages instead of duplicating long definitions in tutorials.
- Use MkDocs admonitions for system behavior, warnings, and explanatory information rather than presenting those explanations as numbered tutorial steps.
- Prefer actual UI labels and exact governed values.
- State when a behavior applies only to the public demo, a Platform Administrator, or a particular application role.
- Update release notes and upgrade guidance when a maintenance change affects behavior or schema.
Tutorial model¶
The Acme Bank tutorial is the canonical from-scratch learning sequence. The Northstar Financial repository is the populated evaluation/demo model. Keep those two purposes distinct.