Content
92%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 well-structured, lean instruction skill with a clear sequenced workflow, validation checkpoints, and clean one-level-deep reference split. The only gap is the absence of a runnable worked example, which leaves actionability just short of maximum.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and assumes Claude's competence — it never explains what a chart or library is, and every line delivers a concrete rule, table, or decision criterion with no padding. | 5 / 5 |
Actionability | Highly actionable instruction-only guidance: a concrete tool-selection table (Observable Plot, Vega-Lite, G2, ECharts, TanStack, etc.), explicit encoding rules ('条形通常从零开始'), and checklists — but lacks a worked runnable example that would make it copy-paste ready. | 4 / 5 |
Workflow Clarity | A clear sequenced pipeline ('读者问题 → 数据语义与状态 → view model → 关系与编码 → 视觉系统 → 代表视图 → 完整状态') with explicit validation checkpoints (representative view must pass before extending, recompute with representative values) and a final completion checklist in section 8. | 5 / 5 |
Progressive Disclosure | Clear overview with two well-signaled one-level-deep references — [领域模式](references/patterns.md) and [工具能力档案](references/tool-profiles.md), both verified to exist — and explicit guidance to load only relevant sections, keeping the SKILL.md body an overview while detail lives in bundles. | 5 / 5 |
Total | 19 / 20 Passed |