Agenthog Experiment

Run an A/B test end-to-end with AgentHog as the source of truth — create a feature flag, wire the variant read into the app, verify exposure is flowing, open an experiment window with the right metric and horizon, read results honestly, ship the winner, and retire the flag. Use when setting up an A/B or A/B/n test, ramping a rollout percentage, checking experiment results, or cleaning up finished flags.

AnniesAI Updated

File contents

AnniesAI/agenthog-claude commit bb65f42cc1

Frequently asked questions

npx get-skill add anniesai/agenthog-experiment