Content
63%Scale 1-5Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid instructional skill that covers the A/B testing workflow comprehensively with useful tables, checklists, and frameworks. Its main weaknesses are moderate verbosity (explaining concepts Claude already understands), dead references to non-existent bundle files, and some sections that could be more tightly structured. The hypothesis framework and sample size quick reference are particularly strong elements.
Suggestions
Create the referenced bundle files (references/sample-size-guide.md and references/test-templates.md) and move the detailed sample size table and documentation template content into them to improve progressive disclosure.
Trim explanations of concepts Claude already knows—remove definitions of statistical significance, what client-side/server-side means, and the detailed explanation of the peeking problem. Replace with brief reminders or constraints.
Condense the 'When to Use' trigger word list into a shorter description; Claude can infer relevance from context without an exhaustive keyword list.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill contains useful reference tables and frameworks, but includes some unnecessary explanations Claude already knows (e.g., explaining what statistical significance means, explaining the peeking problem in detail, describing what client-side vs server-side means). The 'When to Use' trigger list is overly verbose. Several sections could be tightened. | 3 / 5 |
Actionability | Provides concrete frameworks (hypothesis template), specific sample size tables, checklists, and clear metric selection examples. However, it lacks executable code examples for implementation (e.g., no code for setting up a test in PostHog or LaunchDarkly), and some guidance remains at the conceptual level rather than copy-paste ready. For an instruction-only skill this is mostly strong. | 4 / 5 |
Workflow Clarity | The skill presents a clear sequence from hypothesis → design → implementation → running → analysis → documentation, with a pre-launch checklist and analysis checklist serving as validation checkpoints. However, there's no explicit feedback loop for what to do if tracking verification fails or if QA reveals issues—the checklist items are listed but the error recovery path is implicit rather than explicit. | 4 / 5 |
Progressive Disclosure | References to external files like 'references/sample-size-guide.md' and 'references/test-templates.md' are well-signaled, but no bundle files exist to support them, making these dead references. The skill also inlines a substantial amount of content (sample size tables, metrics examples, common mistakes) that could be split into reference files, making the main file longer than necessary. | 3 / 5 |
Total | 14 / 20 Passed |