Content
96%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 exemplary single-file skill: dense domain-specific rules, executable commands and code throughout, a sequenced workflow with validation feedback loops, and a completion checklist. The only structural improvement is offloading the advanced ScreenSpace/HUD placement guidance to a reference file now that the body has grown past overview size.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every line delivers IWSDK-specific knowledge Claude cannot already know ('numeric UIKit sizes are centimeters (100 = 100 cm = 1 m)', 'UIKitML surfaces are single-sided') with zero padding or generic explanation, so every token earns its place. | 5 / 5 |
Actionability | The body provides copy-paste-ready CLI commands with exact JSON arguments ('npx @iwsdk/cli ui render-preview --input-json ...'), complete TS/JSON code samples for manifest registration, scene nodes, and runtime wiring, and named commands for every verification step. | 5 / 5 |
Workflow Clarity | Sections 1-6 form a clear sequence (identity -> editor -> isolated render -> scene placement -> editing -> runtime), each edit cycle has an explicit rerender/inspect/logs feedback loop, and a completion checklist closes the process with concrete validation criteria. | 5 / 5 |
Progressive Disclosure | The file is well-sectioned with a single clearly-signaled external doc pointer ('See docs/concepts/spatial-ui/hud.md'), but no bundle files exist and all ~190 lines live inline in SKILL.md; the advanced ScreenSpace/HUD placement guidance could be split into a reference file, matching 'good structure; most content appropriately placed' rather than the fully-split anchor 5. | 4 / 5 |
Total | 19 / 20 Passed |