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.
A well-structured skill body: environment setup with a conditional install checkpoint, a request-to-method mapping table, terse key rules with concrete limits, and clean deferral to a single real reference file. The main gaps are the duplicated intro sentence, a skeleton general-pattern example that never shows the mandated try/except wrapper, and no explicit error-recovery loop for rate-limited or empty responses.
Suggestions
Delete the duplicated intro sentence under the title (the frontmatter description already covers it) and trim the host-specific shell-probe preamble to one short line.
Show one complete worked example (e.g., a try/except around ticker.history with an empty-DataFrame check) so the 'always wrap in try/except' rule is executable rather than aspirational.
Add a brief error-recovery step in Step 3 for rate limits or empty data (e.g., retry with a delay, then report the failure) to close the feedback loop.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and table-driven with little padding, but has minor trim opportunities: the intro sentence 'Fetches financial and market data from Yahoo Finance using the yfinance Python library' duplicates the frontmatter description, and the Codex/shell-probe preamble adds meta-explanation. This fits the 'efficient; minor instances of over-explanation' anchor, not the lean 5. | 4 / 5 |
Actionability | Concrete executable pieces abound (install probe, install snippet, request-to-method mapping table, period/interval tables, specific intraday limits), but the general pattern is a skeleton ('# ... use the appropriate method from the reference') and rule 1 mandates 'Always wrap in try/except' without ever showing such an example. Mostly executable with minor gaps — anchor 4, not the fully copy-paste-ready 5. | 4 / 5 |
Workflow Clarity | A clear four-step sequence (ensure availability, identify need, write/execute code, present data) with a conditional checkpoint in Step 1 (install if YFINANCE_NOT_INSTALLED). There is no explicit error-recovery loop for rate limits or empty data beyond the try/except mandate. This is read-only fetching, so the destructive/batch cap does not apply; it fits 'clear sequence with most checkpoints present; minor validation gaps' rather than the feedback-loop anchor at 5. | 4 / 5 |
Progressive Disclosure | The body is a lean overview (mapping table, key rules, valid parameters) with the bulk API detail split into a verified one-level-deep reference, clearly signaled in a 'Reference Files' section with purpose and a usage cue ('Read the reference file when you need exact method signatures or edge case handling'). This matches the top anchor for clear overview with well-signaled references. | 5 / 5 |
Total | 17 / 20 Passed |