author-eval

What problem does it solve?

This Skill automates the creation of eval.json for Claude Code skills, streamlining the evaluation process and enhancing coverage.

Core Features & Use Cases

• 2-pass pipeline: Generates an initial local draft with Ollama, then refines with Sonnet.
• Batch processing: Handle multiple skills with options for explicit slugs or random selection.
• Validation: Orchestrates the final verification to ensure parser cleanliness.
• Use Case: Use /author-eval closeweek for weekly batch evaluations, or /author-eval --random 5 to evaluate a random set of 5 skills.

Quick Start

Execute /author-eval closeweek in Claude Code to evaluate the skill for the week.