Ab Testing
When the user wants to plan, run, or read A/B tests, or asks a judgement call on one: is the result real, can we stop…
Skillhawk's judgement
| Measure | Result |
|---|---|
| Grade | D (65/100) |
| Review findings | 3 (1 high, 1 medium, 1 low) |
| Security | No security findings |
| Size | about 3216 tokens, paid on every load |
| Licence | MIT |
| Source | https://github.com/coreyhaines31/marketingskills |
Top review findings
- high
SKILL.md:297[contradiction] contradiction with line 230: "Document every test with" vs "When a test wins, don't just implement it — document the pattern" - medium
SKILL.md:1[bloat] SKILL.md is about 3216 tokens, paid on every load - low
SKILL.md:317[duplicate-rule] same rule as line 113: "do stop tests where guardrails are significantly negative."
Install
npx skillhawk install @community/ab-testing --agent claude
Add -g to install for every project. Original source: coreyhaines31/marketingskills (MIT).