Content
60%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 skill body is well-organized and its CLI documentation is accurate and executable, but the core Python examples describe an API that the bundle does not implement, and every detailed reference file the body points to is missing. The document reads as a spec for a larger toolkit that was never shipped alongside it.
Suggestions
Create the four referenced files (references/bsa_formulas_comparison.md, pediatric_norms.md, chemotherapy_dosing.md, ethnic_adjustments.md) or remove the References section and inline the essential details.
Fix the Python examples to match the actual script: 'from scripts.main import BMIBSACalculator' with calc.calculate(weight_kg, height_cm) and calc.calculate_dose(bsa, dose_mg), or add a scripts/calculator.py implementing the documented API.
State explicitly that only the DuBois BSA formula is implemented, or extend main.py to support the other formulas and pediatric percentile calculation advertised in the body.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is efficient: compact tables for BMI categories, formulas, and CLI parameters, with no padding explaining what BMI is or how Python works. Minor trimming opportunities exist — the 'Key Capabilities' bullet list duplicates the frontmatter description, and the valid ranges (weight 2–300 kg, height 50–250 cm) are stated twice, in 'Input Validation' and again in 'Error Handling'. This matches the anchor 'Efficient; minor instances of over-explanation that could be trimmed', not level 5 where every token earns its place. | 4 / 5 |
Actionability | The CLI section is genuinely executable and matches the bundled main.py (flags --weight/--height/--dose/--format/--output are all real), but the four 'Core Capabilities' Python blocks are not: 'from scripts.calculator import BMIBSACalculator' imports a module that does not exist (scripts/ contains only main.py), and 'calc.calculate_bmi(...)', 'calc.calculate_bsa(..., formulas=[...])', 'calc.calculate_dose(bsa=..., drug=..., dose_per_m2=..., max_dose=...)', and 'calc.pediatric_mode(...)' do not match the actual API (calculate(weight, height) and calculate_dose(bsa, standard_dose_mg)). Only DuBois is implemented, yet Mosteller/Haycock/Gehan-George are presented as working options. This is 'Some concrete guidance but incomplete... missing key details' (level 3) rather than level 4's 'minor gaps' — a large share of the examples fail against the real bundle. | 3 / 5 |
Workflow Clarity | The Workflow section gives a clear 5-step sequence with an explicit validation checkpoint ('Validate inputs are within physiologically plausible ranges; stop if values are outside bounds'), a fallback path with exact response text, a Quick Check smoke command, and a dedicated Error Handling section. It falls short of level 5 because the promised 'manual calculation fallback' is named but never specified, and there is no fix-and-retry loop — matching 'Clear sequence with most checkpoints present; minor validation gaps'. | 4 / 5 |
Progressive Disclosure | Section structure inside SKILL.md is good (overview, workflow, capabilities, CLI, parameters, pitfalls, limitations), but the reference layer is broken: all four files cited under References — 'references/bsa_formulas_comparison.md', 'references/pediatric_norms.md', 'references/chemotherapy_dosing.md', 'references/ethnic_adjustments.md' — do not exist in the bundle (references/ contains only an unreferenced guidelines.md), and the body likewise points to the nonexistent scripts/calculator.py. Scored against the actual bundle per the judging guideline, every signposted path to deeper content dead-ends, which is noticeably below the midpoint — inlining that detail or shipping the files is required to reach level 3. | 2 / 5 |
Total | 13 / 20 Passed |