31 lines
978 B
Markdown
31 lines
978 B
Markdown
# Tests for understanding-konflux-resources
|
|
|
|
This directory contains test scenarios and generated results for validating the skill.
|
|
|
|
## Structure
|
|
|
|
- `scenarios.yaml` - Test scenario definitions with expectations
|
|
- `results/` - Generated test results (one file per scenario sample)
|
|
|
|
## Running Tests
|
|
|
|
**Generate test results** (invokes Claude):
|
|
```bash
|
|
make generate SKILL=understanding-konflux-resources
|
|
```
|
|
|
|
**Run tests** (validates results):
|
|
```bash
|
|
make test SKILL=understanding-konflux-resources
|
|
```
|
|
|
|
## How It Works
|
|
|
|
1. `scenarios.yaml` defines test prompts and expected outcomes
|
|
2. `make generate` invokes Claude with the skill loaded, saves outputs to `results/`
|
|
3. Each result file includes a digest of the skill content on the first line
|
|
4. `make test` validates results match expectations and digest matches current skill
|
|
5. If skill content changes, tests fail until `make generate` is run again
|
|
|
|
See root `Makefile` and `test/` directory for implementation details.
|