checker

I checked out your checker skill and noticed you're tackling input validation without much context—the 60/100 score suggests there's potential here, but developers landing on this will be hunting for a README to understand what makes it different from other validation approaches.

Links:

• GitHub
• Your agentic skill in the SkillzWave marketplace

The TL;DR

You're at 60/100, solidly in D territory. The grading is based on Anthropic's skill architecture best practices—Progressive Disclosure, usability, spec compliance, writing style, and practical utility. Your strongest area is Ease of Use (19/25) with clear metadata and good organization. Your weakest is Writing Style (5/10) where poetic language is crowding out instruction, and Utility (11/20) where validation methods lack concrete frameworks.

What's Working Well

• Metadata is clean — Your description has good trigger terms ("find edge cases," "test assumptions") that help discoverability
• Structure is readable — Headers and sections are clear even in a compact 91-line file; the navigation signals are solid
• Frontmatter is valid — YAML checks out with required fields present and properly formatted

The Big One: Poetic Language Instead of Instruction

Here's what's killing your utility score: sections like "Beautiful Truth" and "Greatest successes = disasters that DIDN'T happen" are philosophy, not guidance. Developers need to use this, not read it.

The fix: Strip the metaphors. Replace "Trust, but verify. Then verify again" with an actual validation checklist—test negative values, zero, infinity, null, empty strings, concurrent access. Create a references/validation-templates.md file with concrete checklists for API validation, UI valida...