ai-agents-skill-testing-after-writing-the-skill-draft-and-before-running-e-81cb710f

THEN Create 2-3 realistic test prompts that a real user would actually say. Share them with the user for confirmation. Save the test cases to `evals/evals.json` with the prompt, expected_output description, and files list. Do not write assertions yet—just the prompts. The assertions will be added later while runs are in progress.

```json
{
  "skill_name": "example-skill",
  "evals": [
    {
      "id": 1,
      "prompt": "User's task prompt",
      "expected_output": "Description of expected result",
      "files": []
    }
  ]
}
```