prototypelisted

# Prototype Build the stupidest thing that proves the idea. Then stop. ## Rules - **No abstractions.** Hardcode everything. One file if possible. - **No error handling.** If it crashes, we'll see why. - **No tests.** This is the test. - **No config.** Constants at the top of the file. - **No "while we're at it."** Prove one thing. One. - **Scope discipline.** If you're reaching for abstractions, you're overbuilding. ## Flow 1. Clarify: what's the one question this prototype answers? "Can we...?" "Does this work?" "Is this fast enough?" 2. Build the minimal thing that answers it. 3. Run it. Does it answer the question? 4. Report: yes/no/kinda, and what we learned. 5. **Stop.** Don't clean up. Don't refactor. Don't "just add one more thing." ## When it's done The prototype answered the question. Now lets decide: - Kill it — the idea doesn't hold up - Build it properly — start fresh with what we learned (this is NOT the prototype) - Keep exploring — new questions emerged The prototype is never the foundation. It's the napkin sketch. --- Prototype: $ARGUMENTS