Content
67%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, highly actionable CLI skill body with clear mode semantics, validated workflow steps, and a genuine error-recovery loop. Its main weaknesses are token bulk (full argument table inline, authoring checklist) and a systematic break in its demo path: every reference to tests/ files points at resources absent from the bundle.
Suggestions
Ship the referenced test bundle (tests/data/expr_matrix.csv, tests/data/group.csv, tests/run_tests.R, tests/test_skill.R) or rewrite the Usage, Testing, and 'When to Read External Files' sections to use user-supplied file placeholders — the canonical example currently fails on first run.
Move the bulk of the 40-row Arguments table into references/cli-guide.md and keep only the frequently-tuned flags (mode, input, groups, method, output_dir) inline to cut tokens.
Delete or relocate the 'Implementation Checklist' and academic 'References' sections; they are authoring-time artifacts and known literature that add context cost without execution value.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is mostly efficient — dense tables, no concept-explainer padding — but it can be tightened: the 40-row Arguments table duplicates detail that references/cli-guide.md exists for, and the authoring-time 'Implementation Checklist' ('get_script_dir() defined before any call to it', 'Documented parameters match script') plus the academic citation list add tokens that do not help execution. Not a 2 because there is no padded explanation of concepts Claude already knows. | 3 / 5 |
Actionability | Commands are copy-paste ready (Rscript scripts/main.R with full flags), input formats are shown as concrete CSV examples, and the Error Handling table gives per-code solutions. The gap: the headline Usage example and Testing section depend on tests/data/expr_matrix.csv, tests/data/group.csv, tests/run_tests.R, and tests/test_skill.R, none of which exist in the bundle, so the canonical command fails out of the box. Still mostly executable rather than high-level, so above anchor 3. | 4 / 5 |
Workflow Clarity | The Workflow section opens with explicit validation (check files, validate columns and matching sample names, validate CLI ranges), states the visualize-mode prerequisite with its error code, and the Error Handling table provides an error-to-fix-to-rerun recovery loop. Anchor 5 is withheld because the documented verification path (tests/run_tests.R and its expected outputs) references files missing from the bundle. | 4 / 5 |
Progressive Disclosure | The 'When to Read External Files' table maps situations to real, one-level-deep references (references/algorithm.md, cli-guide.md, troubleshooting.md, scripts/main.R all verified present), and the Method Selection Guide and error sections point onward clearly. Minor gaps: the body references a bundled tests/data/ directory and test scripts that do not exist, and the full argument table is inlined where cli-guide.md could carry the bulk. | 4 / 5 |
Total | 15 / 20 Passed |