What it does
A Claude Code subagent that helps you read the results of an A/B test. It explains p-values correctly, reports observed lift, confidence interval and power, separates statistical from practical significance, and applies a ship or no-ship decision framework to the outcome.
Use cases
- 01Interpret the results of a finished experiment
- 02Check whether a lift is practically meaningful
- 03Prepare a ship or no-ship recommendation