Free North Star Metric Agent for Claude Code
AI Analyst Lab
Teaching builders to do real product analysis with AI. No SQL, no Python.
Most teams measure the wrong thing. They track fifty metrics, or one vanity number, and none of them say whether the product is actually delivering more value. This open-source agent helps you pick the one metric that does, and figure out what moves it.
What it does
Audits your metric against the 7-question North Star checklist, and refuses vanity metrics like DAU or MRR on the spot, with the reason.
Decomposes what actually drove it over a period, and writes the report for you.
Builds the input tree, and checks each input is a real driver, not your metric renamed.
What you need
Claude Code, and a metric you're unsure about. No SQL, no Python, no setup beyond cloning the repo.
Where it comes from
Every call is grounded in and cited back to Amplitude's North Star Playbook, the framework used by Airbnb, Slack, and Netflix.