Files
tkmind_go/documentation/automation/README.md
T

59 lines
2.0 KiB
Markdown

# Documentation Automation
This directory contains automated pipelines for keeping goose documentation synchronized with code changes.
## Overview
Each automation project tracks specific types of changes and updates corresponding documentation:
| Project | Status | Tracks | Updates |
|---------|--------|--------|---------|
| [recipe-schema-tracking](./recipe-schema-tracking/) | ✅ Active | Recipe schema & validation rules | Recipe Reference Guide |
| cli-command-tracking | 🔮 Planned | CLI commands & options | CLI documentation |
| provider-tracking | 🔮 Planned | Supported AI providers | Provider documentation |
| extension-tracking | 🔮 Planned | Built-in extensions | Extension documentation |
## Architecture
Each automation project follows a consistent pattern:
```
project-name/
├── README.md # Project-specific documentation
├── TESTING.md # How to test this automation
├── config/ # Configuration files
├── scripts/ # Deterministic extraction/diff scripts
└── recipes/ # AI-powered synthesis/update recipes
```
### Design Principles
1. **Modular**: Each project is self-contained
2. **Testable**: Clear inputs/outputs at each stage
3. **Transparent**: Intermediate files can be inspected
4. **Reusable**: Common patterns across projects
### Hybrid Approach
- **Shell scripts**: Deterministic extraction and comparison
- **AI recipes**: Synthesis and documentation updates
## GitHub Actions Integration
Automation projects can be triggered via GitHub Actions workflows in `.github/workflows/`.
See individual project TESTING.md files for workflow usage.
## Adding New Automations
When creating a new automation project:
1. Create a subdirectory: `documentation/automation/your-project/`
2. Follow the standard structure (README, TESTING, config, scripts, recipes)
3. Create corresponding GitHub Actions workflow (if needed)
4. Update this README with the new project
## Questions?
For project-specific questions, see the README in each project directory.