WSWhat Scene?

Learn · Coding agents

What are coding agents still bad at?

As of 2026, coding agents still struggle with big-picture architecture, subtle bugs that only show up at scale, novel problems with no examples to learn from, and knowing when they're wrong. They're confident even when mistaken, so they need a human checking anything important.

They also drift on very large tasks, losing track of earlier decisions, and they can quietly introduce security or performance issues that look fine in isolation. They're pattern-matchers, so genuinely new problems, where there's nothing similar in their training, are where they're weakest.

The practical takeaway is to use them for what they're good at (well-defined, self-contained work with clear success criteria) and to keep human oversight on design, security, and the decisions that carry real consequences.

Updated July 2026

Want this built, not just explained?