📂 Workspace Structure

Here’s how your files will be organized:

<skill>-workspace/
└── iteration-1/
    ├── .lock.json
    ├── eval-1/
    │   ├── with_skill/
    │   │   ├── outputs/
    │   │   ├── timing.json
    │   │   └── grading.json
    │   └── baseline/
    │       ├── outputs/
    │       ├── timing.json
    │       └── grading.json
    └── benchmark.json

🏷️ Running with --models? Each eval gets a model-key subdirectory: eval-1/pi-claude-sonnet/with_skill/....

The .lock.json file 🔒

Every iteration writes a .lock.json file that tracks which evals have finished. It’s what makes --resume work! You don’t need to edit it by hand, but it’s useful to know it exists. If a run is interrupted, the lock stays in "running" status; finish it up with skill-eval run --resume.

🚦 Exit Codes