Anthropic's review philosophy: undo, not reviewers, is the last line of defense for AI-written code
ccerrato147 · x · 2026-10-29
A thread from Addy Osmani (Claude Code team) on reviewing code when Claude writes 90% of it: the last line of defense isn't a reviewer, it's undo — feature flags, staged rollouts, and one-step reverts. Every incident becomes a test the review system is graded on daily.
Data point: from June to September, merged PRs more than doubled, while the share needing a revert fell by over 40%. "Owning" code doesn't mean reading every line — it means three things: fix breakage in minutes with a revert or flag; login, billing, and build/release code always gets a human approver; and before changing a module, ask Claude to explain it to you.
Related event: How Anthropic reviews code when Claude writes 90% of it(6 posts)→
More from coding & agent
- Windows devs can format drives with ReFS to get ZFS-style fast cloning for worktrees — pvncher · 2026-10-29
- NVIDIA Launches CUDA Live: Introduction to CUDA Rust Programming — NVIDIAAI · 2026-10-29
- Developer trains custom game-playing models with his agent, builds Victris Arena — tristanbob · 2026-10-29
- Interface Lineage opens: 15 years of collected interfaces, 500+ patterns free to read — round · 2026-10-29
- Designer runs a daily bot to scout the web for design 'specimens' matching her taste — round · 2026-10-29
- Jason Kneen shares slop_eisley 0.6.22 roadmap for his AI-built FPS game — jasonkneen · 2026-10-29