Content
82%Weight 40%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
An actionable, lean playbook: every section is executable and copy-paste ready, with a clear install-configure-measure flow. Weaknesses are minor — small redundancies between description, intro line, and checklist, and no guidance on diagnosing scores that fall short of the 95 target.
Suggestions
Add a short feedback loop after the Measuring section: what to check first when a Lighthouse category scores below 95 (e.g., failed LCP → verify preload/fetchpriority, render-blocking CSS → verify critters output).
Remove the redundant intro sentence under the h1 (it repeats the frontmatter description verbatim) and trim the closing checklist to items not already explicitly instructed in the body.
If the skill grows beyond this size, move the full Layout.astro pattern into references/layout-pattern.astro and keep a short summary inline.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is code-first and lean — install commands, config, layout pattern, measure — with almost no explanation of concepts Claude already knows. Not 5, because the line under the h1 ("Astro-specific optimizations for 95+ Lighthouse scores.") repeats the frontmatter description verbatim, and the closing checklist restates content already covered in the body. | 4 / 5 |
Actionability | "npm install astro-critters @playform/compress", the complete astro.config.mjs, the compress() options block, the full Layout.astro pattern, and "npx lighthouse ... --preset=perf --form-factor=mobile" are all copy-paste ready and cover the common cases. Nothing is pseudocode or hand-wavy. | 5 / 5 |
Workflow Clarity | There is a coherent implicit sequence (install → configure integrations → apply layout pattern → measure with lighthouse), with the Measuring section and end checklist serving as verification. Not 5, because there is no feedback loop for what to check or fix when a score falls short of 95, and not 3 because the sequence and verification command are clearly laid out. | 4 / 5 |
Progressive Disclosure | Sections are well-organized (Quick Setup, Integrations, Layout Pattern, Measuring, Checklist) with content appropriately placed for a single-file skill of this size; the cross-links to perf-lighthouse and perf-web-optimization are clearly signaled. Not 5, because no bundle structure exists and the ~55-line Layout.astro block is a candidate for a references file if the skill grows, keeping this at good-structure rather than exemplary. | 4 / 5 |
Total | 17 / 20 Passed |