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.
A strong, highly actionable skill whose code is copy-paste ready and covers every effect it names. Its weaknesses are modest: a vocabulary section that re-explains known concepts, no explicit QA-failure feedback loop, and a monolithic single-file layout that could benefit from splitting per-effect code into references.
Suggestions
Cut or drastically shorten the 'Effect Vocabulary' section — definitions like 'Parallax scrolling: background and foreground layers move at different speeds' are concepts Claude already knows; the specific parameter ranges in 'Motion Tokens' are the only part that adds value.
Add a feedback loop to the QA workflow: after 'Run browser QA across desktop and mobile', state what to do when a checklist item fails (fix, re-run ScrollTrigger.refresh(), re-verify) so the sequence has explicit error recovery.
Move the self-contained per-effect implementation blocks (preloader, split text, sticky card stack, parallax) into a references/ file (e.g., references/effects.md) and keep SKILL.md as an overview with one-level-deep pointers, trimming inline token weight.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dominated by dense, useful code and concrete tokens, but the 10-bullet 'Effect Vocabulary' section explains concepts Claude already knows (e.g., 'Parallax scrolling: background and foreground layers move at different speeds') and could be trimmed. This is minor over-explanation rather than pervasive padding, so it sits at anchor 4, not 3. | 4 / 5 |
Actionability | Everything is executable: an install command, a complete Lenis+ScrollTrigger setup, and copy-paste-ready functions for the preloader, split-text reveals, section reveals, scrubbed scenes, the sticky card stack, parallax, and the footer reveal. This matches the anchor for fully executable, copy-paste ready code covering the common cases. | 5 / 5 |
Workflow Clarity | The 'Build Order' section gives a clear 9-step sequence ending with reduced-motion fallbacks and 'Run browser QA across desktop and mobile', and a QA Checklist provides concrete checkpoints. However, there is no explicit validate-then-fix feedback loop for failing QA checks, which keeps it below the anchor-5 example's error-recovery structure. | 4 / 5 |
Progressive Disclosure | Sections are clearly headed and easy to navigate, but this is a ~420-line single file with no bundle files; several self-contained per-effect code blocks (e.g., the sticky card stack and progression scenes) could live in one-level-deep reference files. Structure is good with minor organization gaps rather than misplacement, so it fits anchor 4 rather than 3. | 4 / 5 |
Total | 17 / 20 Passed |