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.
An exceptionally actionable, well-structured instruction skill: every pattern is defined by concrete trigger phrases and demonstrated with a before/after rewrite, and the workflow embeds a genuine self-audit feedback loop. The main improvements are removing the duplicated task/process sections and splitting the pattern catalog into a reference file so SKILL.md stays a lean overview.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The body is dense and example-driven: each of the 24 patterns gets a watch-for phrase list and a tight Before/After pair, with almost no explanation of concepts Claude already knows. It is not a 5 because the "Your Task" and "Process" sections largely duplicate each other (identify, rewrite, audit, present appears twice), and the Invocation code block restates slash-command mechanics, so some tokens could be trimmed. | 4 / 5 |
Actionability | Guidance is fully concrete and copy-paste ready for an instruction-only skill: specific watch-for phrase lists per pattern (e.g. "stands/serves as, is a testament/reminder"), Before/After rewrites for every pattern, a fixed Output Format template (Draft Rewrite / AI Audit / Final Version / Changes Made), and a complete worked example from input to final version. This matches the anchor for fully executable guidance covering the common cases. | 5 / 5 |
Workflow Clarity | The process is an explicit sequence with a built-in validation feedback loop: draft, then "Ask: 'What makes this still obviously AI generated?'", list remaining tells, revise based on the audit, and present the final version. This matches the anchor for clear sequencing with explicit validation and error-recovery loops; the operation is non-destructive, so no batch-operation cap applies. | 5 / 5 |
Progressive Disclosure | Section structure is good (grouped categories, numbered patterns, clear headers), but the entire ~390-line skill is one flat file with no bundle files at all; the 24-pattern catalog (~250 lines of reference material) is inlined in SKILL.md where a split (patterns in references/, process and examples in the body) would match the anchor for content that should be separate sitting inline. It is above anchor 2 because what is present is well organized and easy to navigate, not buried. | 3 / 5 |
Total | 17 / 20 Passed |