Computing Life · Share · YageJul 9, 2026, 08:00 UTC+8

Clean code doesn't boost agent pass rates, but it cuts navigation costs

你不需要为 AI 把代码擦得像样板间，但你需要让它找得到路

A SonarSource paper ran 660 controlled trials with Claude Code + Claude Sonnet 4.6 on clean vs messy code pairs. Pass rates differed by less than 1 percentage point, but clean code cut input tokens by 7.1%, output tokens by 8.5%, and file revisitation by 34%. Multi-module tasks saw input tokens drop 10.7% and revisitation drop 50.8%. HN commenters noted the pairs were auto-generated, not real-world degraded code, and the authors admitted they didn't run full regression tests. The real takeaway: clean code doesn't raise success rates, it lowers the context cost of search, verification, and review. The highest-ROI practices are single sources of truth, removing dead code and stale patterns, explicit module boundaries, and executable lint/test feedback loops for agents.

Why it matters: SonarSource ran 660 controlled trials with Claude Code. Counterintuitive result: clean code didn't improve pass rates, but cut token use by 7-8% and file revisits by 34%. Concrete numbers, clear experimental design, HN discussion as corroboration. HKR all hit. Deduction: tasks...

Read the original ↗Export Markdown

Take

This one's worth opening because it turns a gut feeling into something measurable. SonarSource ran 660 controlled trials with Claude Code + Claude Sonnet 4.6 on clean vs messy code pairs. Pass rates barely moved — less than 1 percentage point difference. But clean code cut input tokens by 7.1%, output tokens by 8.5%, and file revisitation by 34%. Multi-module tasks saw revisitation drop by over 50%.

Don't read this as "cleaner code makes agents smarter." What it actually shows is lower navigation and verification cost. The messy side even had a slightly higher pass rate — the gap is too small to take seriously, but it's a reminder that splitting logic across more helpers and abstraction layers doesn't automatically help an agent doing a short, focused task.

I'd discount the HN skepticism a bit, but not ignore it. The clean/messy pairs were auto-generated, not real-world degraded code. The authors also admitted they didn't run full regression tests after each task — we don't know if clean or messy code caused more collateral damage in untouched modules. Still, the core signal holds: agents search, read files, and trace call chains like humans do. Naming, file location, and module boundaries directly affect how much context they burn.

The highest-ROI moves aren't mysterious: single sources of truth, deleting dead code and stale patterns, explicit module boundaries, and executable lint/test feedback loops. You don't need a showroom codebase. You need one where the agent can find its way without doubling back.

Source: Computing Life · Share · Yage · yage.ai
Published: Jul 9, 2026, 08:00 UTC+8 · Jul 9, 2026