Content
35%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill provides well-structured output templates for three design system tasks but is undermined by excessive verbosity — particularly the 'Components of a Design System' section which explains concepts Claude already knows. The templates themselves are useful but lack accompanying procedural guidance on *how* to perform audits or detect issues, making the skill more of a formatting guide than an actionable workflow.
Suggestions
Remove the entire 'Components of a Design System' section — Claude already knows what design tokens, components, and patterns are. This saves ~20 lines of unnecessary context.
Add concrete procedural steps for the audit workflow: e.g., 'grep for hardcoded hex values (#[0-9a-f]{6})', 'check component file naming against convention X', 'compare prop names across components for consistency'.
Move the three large output templates to separate referenced files (AUDIT_TEMPLATE.md, DOCUMENT_TEMPLATE.md, EXTEND_TEMPLATE.md) and keep only a brief summary of each in the main SKILL.md.
Add a validation step to the audit workflow — e.g., 'After generating the audit, verify findings by spot-checking 2-3 flagged components before presenting results.'
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is very verbose, spending significant tokens on listing design system concepts (tokens, components, patterns) that Claude already knows well. The 'Components of a Design System' section is pure concept explanation that adds no actionable value. The output templates, while useful, are extremely long and could be condensed significantly. | 1 / 3 |
Actionability | The output templates provide concrete structure for what to produce, which is helpful. However, there are no executable code examples, no specific commands or tools to run for auditing, and the guidance is largely template-based rather than instructing Claude on *how* to perform the audit (e.g., what to grep for, how to detect hardcoded values). The templates are filled with placeholders rather than realistic examples. | 2 / 3 |
Workflow Clarity | The three subcommands (audit, document, extend) are clearly separated with distinct output formats, but there's no step-by-step process for how to actually perform each task. For example, the audit workflow doesn't specify: scan files → identify patterns → compare against tokens → generate report. There are no validation checkpoints or feedback loops for any of the three workflows. | 2 / 3 |
Progressive Disclosure | The content is structured with clear sections and references CONNECTORS.md appropriately. However, the three massive output templates make the file very long and monolithic — these could be split into separate reference files (e.g., AUDIT_TEMPLATE.md, DOCUMENT_TEMPLATE.md, EXTEND_TEMPLATE.md) with the main file providing a concise overview. | 2 / 3 |
Total | 7 / 12 Passed |