Content
27%Scale 1-3Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill contains some useful executable code (PostHog integration, cohort retention, A/B testing) but is undermined by significant verbosity from generic boilerplate sections, product-specific content (Auri) that limits reusability, and poor organization. The content would benefit greatly from removing filler sections, splitting detailed code into referenced files, and providing clearer workflow sequencing with validation steps.
Suggestions
Remove all generic boilerplate sections (Best Practices, Common Pitfalls, Limitations, When to Use/Do Not Use) — they contain no domain-specific value and waste tokens.
Split the detailed code examples (cohort retention, A/B test calculator, event taxonomy) into separate referenced files and keep SKILL.md as a concise overview with navigation links.
Add a clear end-to-end workflow that sequences the steps: define events → implement tracking → validate data flowing → build funnels → analyze cohorts → set up dashboards, with explicit validation checkpoints at each stage.
Either generalize the content beyond 'Auri' to be reusable for any product analytics setup, or clearly frame it as Auri-specific configuration with a general-purpose section.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is bloated with generic boilerplate sections (When to Use, Do Not Use, Best Practices, Common Pitfalls, Limitations, Related Skills) that add no value and are vague platitudes. The Deming quote is unnecessary. The content explains concepts Claude already knows and includes product-specific data (Auri events, Auri funnel) that is highly opinionated context rather than reusable instruction. The 'Best Practices' and 'Common Pitfalls' sections are entirely generic filler. | 1 / 3 |
Actionability | The skill provides executable Python code for PostHog tracking, cohort retention calculation, A/B test significance, and feature flags, which is good. However, much of the content is descriptive rather than instructive (e.g., the funnel visualization is a static diagram, the North Star section describes a framework rather than giving executable steps). The commands table lists slash commands with no implementation details. The code is mostly copy-paste ready but tied to a specific product ('Auri') limiting general applicability. | 2 / 3 |
Workflow Clarity | The funnel optimization section provides a 6-step process with validation (p-value < 0.05, 2 weeks minimum), which is decent. However, there's no overall workflow tying the pieces together — the skill reads as a collection of code snippets and frameworks without clear sequencing. The 'How It Works' section is just a naming convention snippet with no actual workflow. Missing validation checkpoints for the implementation steps (e.g., verifying PostHog connection, validating event data). | 2 / 3 |
Progressive Disclosure | The content is a monolithic wall of text with no references to external files despite being over 200 lines. The event taxonomy, cohort analysis code, A/B testing calculator, and benchmarks could all be split into separate reference files. The Related Skills section mentions other skills but provides no meaningful navigation. No bundle files exist to support progressive disclosure. | 1 / 3 |
Total | 6 / 12 Passed |