Content
56%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 has a genuinely actionable core — a concrete CLI invocation with documented PICO arguments and a well-specified output format — wrapped in a large amount of generic template boilerplate. Cutting the duplicated validation sections, the verbatim description repeat, and the platitudinous workflow/safety sections would roughly halve the body while losing nothing.
Suggestions
Delete or merge the redundant generic sections (When to Use, Key Features, Dependencies, Implementation Details, Required Inputs, Recommended Workflow, Output Contract, Validation and Safety Rules, Deterministic Output Rules, Completion Checklist) — keep only the concrete ones: the CLI usage, Arguments, Output Format, and a single validation step.
Consolidate "Validation Shortcut" and "Quick Validation" into one section with the single command "python scripts/main.py --help" and its expected output.
Remove the verbatim repetition of the frontmatter description inside "Key Features" and drop the odd dated example path ("cd "20260316/scientific-skills/...") in favor of a portable invocation example.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Roughly half the body is generic boilerplate that adds nothing Claude cannot infer: "Use this skill when the request matches its documented task boundary" is tautological, the frontmatter description is repeated verbatim in Key Features, and the same "python scripts/main.py --help" validation appears in both "Validation Shortcut" and "Quick Validation". This matches anchor 2 (noticeably verbose, several padded sections), not anchor 3, because the padding is pervasive rather than occasional. | 2 / 5 |
Actionability | The core path is executable and copy-paste ready: "python scripts/main.py --type efficacy --P "Type 2 Diabetes" --I "Metformin" --C "Placebo" --O "HbA1c" --study_design "RCT"" with documented arguments and a described output format (headers, "[LLM Instruction]" blockquotes, hard data). It is anchor 4 rather than 5 because no output path/flag is documented and the surrounding sections ("Recommended Workflow", "Output Contract") stay abstract. | 4 / 5 |
Workflow Clarity | "Usage for Agents" gives a clear sequence (call the skill with the command → read the output file → execute its instructions), and validation checkpoints exist (Quick Validation expected output, Completion Checklist, Failure Handling recovery steps). It falls short of anchor 5 because validation is scattered across redundant sections rather than integrated as an explicit validate→fix→retry loop within the sequence. | 4 / 5 |
Progressive Disclosure | No bundle files exist (references/, scripts/, assets/ are absent), so this scores the body structure alone: it has clear section headers and mentions script paths ("scripts/main.py", "scripts/calculators/sample_size.py"), but it is a 20+ section single file with substantial inline boilerplate and duplicated content that should be consolidated. This matches anchor 3 (some structure but could be better organized), not anchor 4 (the duplication and template filler are real organization gaps). | 3 / 5 |
Total | 13 / 20 Passed |