Read Mode
What does this code do?
Output, return values, bug spotting, and safest-pattern questions for AI-flavored snippets.
Python fluency for AI builders
Patch tiny failures with confidence.
Train on compact Python reps pulled from the world of model outputs, eval loops, tool calls, structured data, and guardrails. No LeetCode. No setup pain. Just practical short-session reps.
2-10 minute reps built for short sessions and repeat practice.
Pick your practice slice
Read Mode
Output, return values, bug spotting, and safest-pattern questions for AI-flavored snippets.
Debug Mode
Short debugging reps focused on syntax slips, wrong conditions, parsing mistakes, and loop bugs.
Patch Mode
Minimal patches for safer access, validation, refusal handling, and malformed output protection.
Jump into the next available challenge using your filters.
Your dojo board
Clears XP, streaks, and challenge history stored in your browser on this device.
How it works
Short Python reps for AI builders working with model outputs, eval rows, tool responses, and guardrails.
Patch questions use lightweight matching with trimmed whitespace and a small set of accepted answers when needed.
Use keys 1-4 for multiple choice answers, Enter to submit, and N for next challenge after feedback.