Content
85%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
The body is a tight, well-structured strategy spec with clear signal logic and useful pitfall notes. Its main gap is actionability: it gives formula fragments rather than a complete, executable implementation.
Suggestions
Provide one complete, copy-paste-ready pandas function that computes the signal end-to-end (including the fillna handling noted in Pitfalls) instead of inline formula fragments.
Show a short runnable example with sample OHLCV input and expected signal output so the convention (1/-1/0) is verifiable.
Optionally note how the signal feeds position sizing or backtesting to clarify the boundary of this skill's responsibility.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is lean and efficient, assuming Claude's knowledge of pandas and finance; formulae are given inline (e.g. 'returns.rolling(hv_window).std() * sqrt(252)') with no padded explanations, so every token earns its place. | 3 / 3 |
Actionability | It provides concrete formulae and a parameters table with defaults, but these are code fragments rather than a complete, copy-paste-ready executable block, leaving key implementation details (full signal function, fillna handling) implicit. | 2 / 3 |
Workflow Clarity | Signal Logic is a clear 3-step sequence (compute HV, percentile-rank, generate signal) with threshold rules and an edge-case note (fillna before lookback fills); as a simple single-purpose skill the process is unambiguous, qualifying for the top score. | 3 / 3 |
Progressive Disclosure | The skill is under 50 lines with no need for external references, and its well-organized sections (Purpose, Signal Logic, Key Implementation Details, Parameters, Pitfalls, Dependencies, Signal Convention) satisfy the simple-skill top score. | 3 / 3 |
Total | 11 / 12 Passed |