Content
81%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 well-structured, actionable workflow with strong validation checkpoints and lean prose. Minor trimming of editorial annotations and pulling the prompt template into a reference file would push it to a 5.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean — tool signatures, field definitions, and bullet lists — and assumes Claude knows Sharpe/PnL/equity curve without explaining them; minor editorial bits like '(PDF Section 5 — gut punch)' could be trimmed, so just below a 5. | 4 / 5 |
Actionability | Provides concrete tool calls with parameters and return values (extract_shadow_strategy(journal_path=...), run_shadow_backtest(...)) plus a copy-paste rule-translation prompt template; the internal extractor._translate_rule fallback is referenced but not shown, a minor gap. | 4 / 5 |
Workflow Clarity | A clear four-step sequence with explicit validation/feedback loops — confirm rules then re-run with higher min_support, raise on profitable roundtrips < 5, and weasyprint → HTML-only fallback — plus a 红线 constraints checklist. | 5 / 5 |
Progressive Disclosure | Well-organized into clear sections (何时触发, 工作流, 产出解读, 对话模板, 规则翻译 Prompt 模板, 红线) and self-contained with no external references; at ~70 lines it slightly exceeds the under-50 simple-skill threshold, so just below a 5. | 4 / 5 |
Total | 17 / 20 Passed |