Stumbled across your skill while looking at the latest submissions—the approach here is solid, scoring an 86. One thing I'm curious about: what made you structure it this way instead of going with the more conventional pattern most skills use?
Links:
The TL;DR
You're at 86/100, solid B territory. This is based on Anthropic's skill best practices. Your strongest area by far is Writing Style (9/10)—clean, imperative tone throughout. The weakest spot is Spec Compliance (12/15) where you're missing some trigger phrase coverage, but that's a quick fix.
What's Working Well
- Layered structure is chef's kiss: SKILL.md stays lean with the heavy lifting delegated to
references/syntax.md and references/architecture.md. That's exactly what Progressive Disclosure Architecture should look like.
- Writing is consistently solid: No marketing fluff, no second-person rambling. Every sentence does work. "Use when writing, modifying, debugging, or understanding Lea code" tells me exactly what this is for.
- Reference files are well-organized: Two reference files, one level deep, exactly the right depth. Shows you understand the layering principle.
- Utility is legit: This fills a real gap for a niche language. The syntax examples and quick start actually teach something.
The Big One
syntax.md is 361 lines with no table of contents. When someone drops into a reference file that long, they're lost. You've got sections for Comments, Bindings, Functions, etc., but they can't scan to find what they need.
Add a TOC right after the title:
# Lea Syntax Reference
## Contents
- [Comments](#comments)
- [Bindings](#bindings)
- [Functions](#functions)
- [Pipes](#pipes)
- [Decorators](#decorators)
This alone gets you +3 points. It's a 2-minute fix that massively improves navigation.
Other Things Worth Fixing
-
Generic description trigger – You list "writing or modifying Lea code" but the skill clearly handles debugging and understanding too. Expand it: "Use when writing, modifying, debugging, or understanding Lea code." That's another trigger phrase and gets you closer to full coverage. +1 point.
-
Validation workflow gap – Your "Running" section shows commands (npm run repl, npm run lea file.lea) but nothing on how to validate if the code actually works. Add a quick validation checklist: check syntax errors, verify outputs with print statements. +1 point.
-
Some redundancy in syntax examples – Multi-line records show the same pattern twice. Consolidate to one canonical example per concept. Tightens up the writing style score. +1 point.
Quick Wins
- Add TOC to syntax.md (biggest impact)
- Expand description triggers to include debug/understand use cases
- Add validation guidance to the Running section
- Consolidate redundant examples in syntax reference
Checkout your skill here: [SkillzWave.ai](https://skillzwave.ai) | [SpillWave](https://spillwave.com) We have an agentic skill installer that install skills in 14+ coding agent platforms. Check out this guide on how to improve your agentic skills.
Stumbled across your skill while looking at the latest submissions—the approach here is solid, scoring an 86. One thing I'm curious about: what made you structure it this way instead of going with the more conventional pattern most skills use?
Links:
The TL;DR
You're at 86/100, solid B territory. This is based on Anthropic's skill best practices. Your strongest area by far is Writing Style (9/10)—clean, imperative tone throughout. The weakest spot is Spec Compliance (12/15) where you're missing some trigger phrase coverage, but that's a quick fix.
What's Working Well
references/syntax.mdandreferences/architecture.md. That's exactly what Progressive Disclosure Architecture should look like.The Big One
syntax.md is 361 lines with no table of contents. When someone drops into a reference file that long, they're lost. You've got sections for Comments, Bindings, Functions, etc., but they can't scan to find what they need.
Add a TOC right after the title:
This alone gets you +3 points. It's a 2-minute fix that massively improves navigation.
Other Things Worth Fixing
Generic description trigger – You list "writing or modifying Lea code" but the skill clearly handles debugging and understanding too. Expand it: "Use when writing, modifying, debugging, or understanding Lea code." That's another trigger phrase and gets you closer to full coverage. +1 point.
Validation workflow gap – Your "Running" section shows commands (
npm run repl,npm run lea file.lea) but nothing on how to validate if the code actually works. Add a quick validation checklist: check syntax errors, verify outputs with print statements. +1 point.Some redundancy in syntax examples – Multi-line records show the same pattern twice. Consolidate to one canonical example per concept. Tightens up the writing style score. +1 point.
Quick Wins
Checkout your skill here: [SkillzWave.ai](https://skillzwave.ai) | [SpillWave](https://spillwave.com) We have an agentic skill installer that install skills in 14+ coding agent platforms. Check out this guide on how to improve your agentic skills.