evals.json

{
  "skill": "ocas-forge",
  "version": "1.0.0",
  "description": "Evaluation test cases for ocas-forge",
  "evals": [
    {
      "input": "Design a new skill for managing Docker containers",
      "expected_output": "Creates a skill package with proper SKILL.md frontmatter, command definitions, and reference files",
      "description": "Skill design and construction"
    },
    {
      "input": "Validate an existing skill package",
      "expected_output": "Runs validation checks on structure, frontmatter, and content quality",
      "description": "Skill validation"
    },
    {
      "input": "Build a skill from a specification document",
      "expected_output": "Produces a complete skill package ready for testing",
      "description": "Skill building from spec"
    }
  ]
}