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 a concrete, actionable sector-rotation framework with real executable code and explicit numeric scoring rules. It is dense and efficient, though monolithic with no progressive disclosure into reference files.
Suggestions
Split the detailed scoring rubric and ETF code tables into a references/ file (e.g. references/scoring.md) and link from the overview to enable progressive disclosure.
Add an explicit validation/sanity-check step before emitting 超配/低配 recommendations (e.g. verify data freshness and cross-check 动量 vs 景气度 signals).
Trim the three full 产业链 examples to one canonical example plus a one-line summary of the others to reduce token weight.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Dense tabular content that assumes Claude's knowledge of PE/PB/ROE/PMI with minimal padding; a few sections (three detailed 产业链 examples, repeated 周期股估值陷阱 explanation) could be trimmed. | 4 / 5 |
Actionability | Provides an executable pandas sector_momentum function, concrete scoring rules with numeric thresholds, explicit composite-formula weights, specific ETF codes, and a copy-paste-ready output template. | 5 / 5 |
Workflow Clarity | Clear sequenced flow from 景气度评分 → 动量排名 → 产业链传导 → 多维比较 → 配置建议 with explicit decision rules (Top 5 超配/Bottom 5 低配); checkpoints are implicit rather than explicit, though no destructive operation requires hard validation. | 4 / 5 |
Progressive Disclosure | Well-sectioned with clear headers but the entire ~190-line framework is inlined in SKILL.md with no external reference files to split out detailed tables (e.g. the scoring rubric or ETF code list). | 3 / 5 |
Total | 16 / 20 Passed |