Content
50%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A thorough, well-organized reference framework for US ETF flow analysis with concrete code and detailed tables. It is held back by explanatory padding, code examples with undefined placeholders, an implicit rather than explicit workflow, and a monolithic structure that could be split across reference files.
Suggestions
Trim concept explanations Claude already knows (creation/redemption mechanics, price-vs-flow distinction) to tighten conciseness.
Make code examples fully executable by defining the placeholder inputs (spy_flow, daily_flows, aum) or wrapping them in runnable functions.
Move the large ETF reference tables (broad market, sector, style, fixed income, thematic) into separate reference files and keep SKILL.md as a concise overview with clearly signaled links.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient reference material, but spends tokens explaining concepts Claude already knows (creation/redemption mechanics, 'ETF price movement ≠ ETF flow') and could be tightened. It is not a verbose score-1 explainer, but not lean enough for 3. | 2 / 3 |
Actionability | Provides concrete code blocks and tables, but several examples rely on undefined variables (spy_flow, iwm_flow, daily_flows, aum) so they are illustrative rather than copy-paste ready, matching the incomplete score-2 anchor. | 2 / 3 |
Workflow Clarity | The Output Format template and section ordering imply a sequence (broad market → sector → style → fixed income → thematic → composite), but no explicit step list or validation checkpoints are given; the analytical context makes a score-1 too harsh but keeps it at 2. | 2 / 3 |
Progressive Disclosure | Sections are well organized, but the ~270-line body is monolithic with large ETF reference tables inlined that would benefit from being split into separate reference files; no external references are used, so it does not reach the well-signaled one-level-deep score-3 anchor. | 2 / 3 |
Total | 8 / 12 Passed |