Content
85%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-engineered skill body: clear routing into five sub-skills, executable code where it matters most, strong validation and error-visibility practices, and exemplary use of two one-level-deep reference files. The main gaps are minor verbosity (a redundant opening paragraph and a couple of editorial sentences) and prose-only formulas in Sub-Skills D and E where small code skeletons would make every path executable.
Suggestions
Remove the duplicated opening sentence (line 19 repeats the frontmatter description) and trim or move the 'Why this matters' paragraph and A2 editorial sentences to the reference file — they explain concepts Claude already knows or add narration without operational value.
Add short executable code for Sub-Skill D's metric computations (annualized vol, dollar volume, drawdown from history) and at least a function skeleton for E2's chain-wide GEX aggregation, so all sub-skills match the copy-paste readiness of A and B.
E1's NAV-proxy and E3's dealer-share formulas are given only in text blocks; wrapping them in the same small reusable Python snippet (or pointing to one in the reference file) would close the last actionability gaps.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The routing table, code blocks, and category-context tables are tight and mostly non-derivable knowledge, but there is minor padding: the opening paragraph repeats the frontmatter description almost verbatim ('Calculates the premium or discount of an ETF's market price relative to its Net Asset Value'), the 'Why this matters' paragraph explains a concept Claude largely already knows, and A2 includes editorial sentences like 'This helps the user see whether the premium is unique to their ETF...'. | 4 / 5 |
Actionability | Sub-Skills A and B provide complete, copy-paste-ready Python with error handling, and E3/E1 give exact formulas in text blocks. However, Sub-Skill D (volatility, dollar volume, drawdown metrics) and E2's GEX chain aggregation are described only in prose with formulas deferred to the reference file, so guidance is 'concrete code or commands with minor gaps' rather than fully executable throughout. | 4 / 5 |
Workflow Clarity | The sequence is explicit and well-validated: a DEPS probe with an install fallback (a genuine feedback loop), a routing table with defaults, quoteType verification, per-ticker try/except with visible failure counts ('Keep failed or missing-NAV counts visible instead of silently treating them as zero'), stale-NAV caveats, and a Step 3 'Always include / Always caveat' checklist for the response itself. | 5 / 5 |
Progressive Disclosure | Deep technical material (GEX derivations, both dealer-positioning conventions, worked example, ETF universe, category benchmarks) is correctly split into two real, one-level-deep reference files, both verified to exist and accurately described in a dedicated Reference Files section with explicit read-order guidance ('Read this **before** running Sub-Skill E'). The 400-line body holds only routing, core operational code, and interpretation thresholds, and navigation is easy via the routing table. | 5 / 5 |
Total | 18 / 20 Passed |