Oak Parallax Product Experiment

---
name: oak-parallax-product-experiment
description: Designs, audits, analyses, and learns from controlled experiments for digital products and services, including A/B/n and split tests, feature experiments, holdouts, switchbacks, ramps, and quasi-experiments. Invoke for assignment and exposure, sample-ratio mismatch (SRM), metrics and guardrails, minimum detectable effect, power or precision, peeking or sequential inference, interference, novelty, seasonality, heterogeneity, accessibility, harms, rollout, rollback, and long-term product outcomes. Also invoke before shipping from an experiment result. Do not invoke for ordinary feature delivery, usability research without a causal intervention, descriptive analytics, or offline model evaluation without an assignment/exposure question. Compose with parallax-design-experiment for novel, high-stakes, cluster, adaptive, or statistically complex designs.
compatibility: "Requires Python 3.10+ only for the optional structural validator; use the documented manual validation fallback when Python is unavailable."
metadata:
  owned: "true"
  version: "0.1.0"
  collection: "parallax"
---

# Parallax Product Experiment (Claude Code)

Read and follow `.agent/skills/cognition/parallax-product-experiment/SKILL-CANONICAL.md`.