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 instruction-style skill body: dense with repo-specific rules, executable code for the common cases, an explicit priority decision tree, and a verification checklist with feedback loops. Its only structural limitation is that all recipes and templates live inline in one long file rather than being split across reference files.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and assumes Claude's competence: it never explains what MSW, Vitest, React Query, or jsdom are. Every section carries project-specific contract ('Drive the real transport, mock the network', the What NOT/What TO lists, the priority order) — no padded or generic filler anywhere. | 5 / 5 |
Actionability | Fully executable throughout: a copy-paste-ready Vitest test skeleton, a typed fixture file, four recipes with complete code (promise gate, vi.fn handler wrappers, query-string variation), real infrastructure paths ('packages/playground/src/test/msw-server.ts'), and a concrete verification command ('pnpm --filter ./packages/playground typecheck'). | 5 / 5 |
Workflow Clarity | A clear decision sequence ('choose in this order: 1. MSW... 2. Playwright... 3. Pure unit tests'), an explicit Verification Checklist with validation checkpoints (typecheck passes, runs in isolation and in suite without onUnhandledRequest failures), and a feedback loop ('that's a real test failure — fix the fixture, never `as any` it'). | 5 / 5 |
Progressive Disclosure | No bundle files exist (references/, scripts/, assets/ are absent), and the single file is well-sectioned with clear headers making it easy to navigate. However, at ~277 lines with four inline recipes and two full code templates, some material could be split into reference files, so it stops short of the ideally-split anchor 5. | 4 / 5 |
Total | 19 / 20 Passed |