Failure Library
Failure cases
Bad assumptions, weak edits, and lessons worth remembering.
Over-editingDo not reward broad patches built on shallow repository evidence.
Failure case: over-editing after only partial repository understanding
A common failure mode where the agent reads just enough of a repository to sound credible, then expands the patch beyond what the evidence supports.
CodexClaude Code
April 4, 2026Read analysis
Shallow reviewReview quality should be judged by meaningful risk discovery, not by polished volume.
Failure case: confident review that missed the runtime-changing risk
A review can sound sharp, cover style issues, and still miss the one behavior-changing problem that actually matters.
CodexClaude Code
April 3, 2026Read analysis