Content
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a strong, highly actionable skill that provides concrete frameworks, scoring formulas, and detailed examples with realistic data. Its main weakness is length—at this size, some content (frameworks, source guidance tables, examples) could be split into referenced files for better progressive disclosure. The workflow is well-structured with implicit validation checkpoints, and the examples effectively demonstrate what good output looks like.
Suggestions
Split the Frameworks & Best Practices section and Examples into separate referenced files (e.g., FRAMEWORKS.md, EXAMPLES.md) to reduce the main skill's token footprint and improve progressive disclosure.
Remove the duplicate explanation of the Opportunity Score formula—it appears in both the Workflow step 5 and the Frameworks section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is fairly comprehensive but includes some unnecessary elaboration. The 'When to Use' section is somewhat verbose listing every possible trigger phrase. The frameworks section, while valuable, could be tightened—e.g., the Opportunity Score formula is repeated twice. However, most content earns its place and provides genuinely useful guidance Claude wouldn't inherently know (like the source-specific guidance table and signal vs. noise rules). | 2 / 3 |
Actionability | The skill provides highly concrete, actionable guidance: a specific scoring formula (Opportunity Score = Importance × (1 - Satisfaction)), a complete output template with exact table structures, detailed examples showing what good output looks like with specific numbers and reasoning, and clear criteria for each workflow step. The examples are particularly strong with concrete ARR figures, vote counts, and specific recommended actions. | 3 / 3 |
Workflow Clarity | The 8-step workflow is clearly sequenced with logical progression from goal understanding through data collection, categorization, scoring, prioritization, deep-dives, and presentation. Validation is embedded throughout: step 1 confirms the goal before analysis, step 4 checks strategic alignment before scoring, and the 'Signal vs. Noise Rules' serve as validation checkpoints against common analytical errors. The framework includes explicit feedback loops like recency bias checks and the requirement for 3+ independent mentions. | 3 / 3 |
Progressive Disclosure | The skill references related skills at the end (user-research-synthesis, churn-analysis, prd-writing) which is good navigation. However, the content is quite long and monolithic—the frameworks section, source-specific guidance, common mistakes, and examples could potentially be split into referenced files. The output template alone is substantial. For a skill of this length (~150+ lines), more content splitting would improve discoverability. | 2 / 3 |
Total | 10 / 12 Passed |