We’ve all been there. You find a code snippet online that almost works — but it has bugs, security holes, or bad practices. The idea is simple: don’t just flag it as wrong. Correct it constructively.
It might fail silently (producing wrong outputs), fail loudly (crashing with an obscure exception), or—most dangerously—introduce a subtle security vulnerability or a crippling performance bottleneck. You are left staring at a screen wondering: Is the AI wrong, or am I misreading this? idea correct the problematic code answers
"Write a JavaScript function to retry a failed API call up to 3 times with exponential backoff." We’ve all been there
def get_average(nums): sum = 0 for i in range(len(nums)): sum += nums[i] return sum / len(nums) # Fails if nums is empty Correct it constructively
Here’s a framework for fixing problematic code answers:
Sometimes the “corrected” version can still be improved: