Claude Plugin · Open Source
We write PRDs with blind spots we can't see — because our assumptions come from our own context. This plugin surfaces what you didn't know to question, tested against 269 Lenny Rachitsky podcast episodes from PMs at Stripe, Figma, Airtable, Linear, Duolingo, and more.
If you leave your email, I may reach out to ask what you thought. Nothing else.
How it works
Attach a document, paste a section, or continue in a thread where you've been writing. Claude reads the full PRD for context — not just the section you want tested.
Type test assumptions in my GTM section or test assumptions in my metrics section. Works on GTM, metrics, pricing, MVP scope, problem statement, activation, and retention.
Claude extracts implicit assumptions, queries 269 Lenny episodes for relevant guest experiences, and returns verdicts with citations — plus one hard question for your next stakeholder review.
Real output — Onboarding automation PRD, metrics section
Assumption 1
"First transaction completed = an activated merchant who will stay."
Lauryn Isford (Head of Growth, Airtable) — activation rates in the 5–15% range correlate more strongly with long-term retention than higher rates. A merchant who transacts once and churns isn't activated. They're a false positive. Episode: "Mastering onboarding"
⟶ ChallengesAssumption 2
"Reducing onboarding TAT from 10 days to 3–4 days will meaningfully improve activation rates."
Kristen Berman (Irrational Labs) — when users drop out of long onboarding flows, landing them back where they left off dramatically improved completion rates more than raw speed. Continuity matters more than time. Episode: "Using behavioural science to improve your product"
⟶ Validates with nuanceThe hardest question this surfaces
"Can you name merchants who completed first transaction in your last pilot and tell me whether they're still transacting 60 days later? If you don't know, your activation metric may be measuring signup theatre, not real business outcomes."
PRD sections you can stress test