Every web application is a state machine.
Boozang users model application behavior - modules, operations, interdependencies. Not click sequences. All documentation must reflect this paradigm. See .agent/context/testing-philosophy.md.
For AI Agents: Read
.agent/instructions.mdfor complete guidelines.
npm run start # Dev server
npm run build # Production build- Use consistent terminology (see
chatgpt/boozang_dictionary.csv) - Test changes locally before committing
- Update LLM content after significant changes
When editing documentation:
- Keep language clear and concise
- Use consistent terminology (see chatgpt/boozang_dictionary.csv)
- Preserve existing formatting and structure
- Test links work correctly
When working with the chatgpt/ folder:
- Run
node chatgpt/clean-en.mjsafter changes to regenerate cleaned content - Output goes to
chatgpt/pages/ all-crawled.mdcontains combined content
- Test locally:
npm run start - Build:
npm run build - Verify:
npm run serve
| Path | Purpose |
|---|---|
docs/**/*.md |
Source documentation |
chatgpt/pages/*.md |
Processed LLM content |
sidebars.js |
Navigation structure |
docusaurus.config.js |
Site configuration |