Content
78%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.
The content is concise and largely actionable with concrete commands and clear rules. Its main weakness is progressive disclosure: it references bundle files that are not actually present, leaving the documented navigation unresolvable.
Suggestions
Either add the referenced files (react-functional-only.md, async-await-promises.md, no-legacy.md, and the rules/ directory) or remove the references so the Quick Reference table points only to existing material.
Add an explicit validation step after `bun x ultracite fix` (e.g., re-run `ultracite check` to confirm no remaining issues) to strengthen the workflow feedback loop.
Clarify what the `rules/` directory contains or inline the few most important rules so the skill stands on its own.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence — commands, a quick-reference table, and console-logging rules with no padded explanations of what Biome or TypeScript is. | 5 / 5 |
Actionability | It provides copy-paste ready commands (`bun x ultracite fix`) and concrete rules, but the Quick Reference table points to files (react-functional-only.md, async-await-promises.md, no-legacy.md) that are not present in the bundle, leaving minor gaps. | 4 / 5 |
Workflow Clarity | For a simple single-purpose skill the action is unambiguous and the Commands/Quick Fix sections sequence the steps clearly; `ultracite doctor` serves as a diagnostic checkpoint, though explicit validation of the fix is only implied. | 4 / 5 |
Progressive Disclosure | Structure is reasonable with a Quick Reference table pointing to separate rule files, but those referenced files (`rules/`, `react-functional-only.md`, etc.) do not exist in the bundle, so the references are not real and navigation is broken. | 3 / 5 |
Total | 16 / 20 Passed |