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 body is well-structured, lean, and gives genuinely executable guidance (real file paths, config keys, and a docs URL pattern instead of guesses). The main gaps are the absence of a sample import for the skill's central task and no explicit recovery step when the post-regen typecheck fails, plus slight redundancy between the opening sections and the frontmatter description.
Suggestions
Add one copy-paste example import (e.g., importing a client or hook from the barrel index vs. a direct file) to make the skill's central task fully executable.
Remove or merge the "When to Use" and "What It Does" sections, which restate the frontmatter description and add tokens without new instruction.
Extend the typecheck rule with an explicit recovery step: if the typecheck fails after regenerating, fix the affected call sites (or the spec) and regenerate/typecheck again.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes competence (no explanation of what OpenAPI or codegen is), but the "What It Does" section ("Explains where generated code lands", "Shows how to discover the real exports") merely restates the frontmatter description and the "When to Use" bullets duplicate it too — trimmable padding that keeps it at anchor 4 rather than anchor 5's every-token-earns-its-place. | 4 / 5 |
Actionability | Concrete, executable guidance throughout: "Read `kubb.config.ts` for the top-level `output.path`", the exact plugin path "`node_modules/@kubb/plugin-<name>/src/types.ts`", the docs URL pattern, and "Import from a folder's `index.ts` when `output.barrel` is set". It stops short of anchor 5 because the core task — importing and calling generated code — has no copy-paste example (e.g., a sample import of a client or hook). | 4 / 5 |
Workflow Clarity | "Find what was generated" provides a clear numbered 3-step sequence, and the Rules include an explicit validation checkpoint ("Typecheck after regenerating so a spec change that broke a call site shows up") for the regen-rewrite batch operation. It is below anchor 5 only because there is no explicit feedback loop stating what to do when the typecheck fails. | 4 / 5 |
Progressive Disclosure | A ~50-line, single-purpose skill with well-organized sections (Where output lands, Find what was generated, Rules, Related Skills) and no need for external reference files — matching the simple-skill exception where organization alone merits anchor 5. The only link, ../config/SKILL.md, is a one-level sibling-skill pointer, not a nested reference. | 5 / 5 |
Total | 17 / 20 Passed |