Real production bugs. You write the prompt. Hidden tests judge the code.
380+ users from around the globealready competing.
Your startup just launched a shipping feature. Your cofounder wrote the pricing logic over a weekend, and it works — but it's 7 levels of nested if/else statements deep...
def calculate_shipping(weight, zone, membership, express):
if weight > 0:
if zone in ["domestic", "international"]:
if membership in [...]:
if express:
if zone == "domestic":
if weight <= 5:
base = 5.99
...No sign-up required to start
5 categories, 3 difficulty levels. Each tests a different AI collaboration skill.
Craft instructions for Claude. Be specific, be clever. It generates Python code from your prompt.
Code runs against hidden tests. Scored on correctness, efficiency, performance, and quality.
Every problem you solve earns rating points. How high can you go?
Refactor messy code into clean, maintainable solutions.
Make brute-force code lightning fast with smart prompts.
Find and fix subtle bugs — guide AI to spot the issues.
Add features without breaking existing tests.
Build complete features from a requirements spec.