Laughable

Most humor advice is vibes: 'be funny', 'read the room'. Puro's take is the opposite: humor has a formula, and a formula can be taught and drilled. A joke is an anticlimactic progression, with a relatable end, presented with the element of surprise: three ingredients, each a dial. Laughable is the side quest that turns that framework into a working app: learn the formula, then practice it until the pattern recognition is yours.

1. 01

   A framework is only useful if it's drilled, not just read. The app had to move people from understanding the formula to actually applying it under a little pressure.

2. 02

   The AI coach had to grade against Puro's canonical framework (the three ingredients, the relevance-pool zones) and never drift into generic, paraphrased humor theory.

3. 03

   Practice has to be varied or it gets stale fast. One drill mode isn't enough to build real range.

4. 04

   It's a side quest, not a flagship; it had to be genuinely good without becoming a project that ate the roadmap.

The app splits into learn and practice. The learn flow is seven steps (the formula, the three ingredients, the relevance pool, layout, delivery, confidence, beliefs) building the mental model piece by piece. Practice is four modes: Domain Mash, Beer Banters, Question Conversation, and Diagnose, each drilling a different muscle. Behind it sits an AI coaching engine prompted hard against Puro's canonical framework: it scores attempts on anticlimax, relatable end, and surprise, and verdicts them on the relevance-pool zones, never substituting generic humor theory. Progress tracks in localStorage; the whole thing ships on Vercel.

Live at laughable.vercel.app. It's a small, complete thing: a framework turned into a learn-then-drill app with a coach that holds the line on the methodology. As a side quest it does its job: it proves the framework teaches, and it's a clean demonstration of how fast a focused, opinionated product comes together under the Puro AI Labs umbrella.