--- id: "nr-agent-written-code-moves-the-bottleneck-to-review-and-the-systems-arou-1300e6" slug: "agent-written-code-moves-the-bottleneck-to-review-and-the-systems-arou-1300e6" type: "raw_signal" title: "Agent-written code moves the bottleneck to review and the systems around it" description: "Anthropic says the jump in merged changes forced it to build a new source-control system." observed_at: "2026-10-21T13:22:14.659Z" record_date: "2026-10-21" date_kind: "observed_at" novelty: "notable" source_platform: "newruntime.com" topics: ["anthropic","claude"] entities: ["Anthropic","Scion","Git","Claude","Antique","Simon Willison","Spanner","Claude Mythos Preview","Diff Tour","Claude Code"] related_patterns: ["harness-architecture-outlives-model-choice","verification-bandwidth-is-the-scarce-resource","ai-native-orgs-move-to-review-and-orchestration"] source_url: "https://simonwillison.net/2026/Oct/13/redesigning-source-control-at-anthropic" source_urls: ["https://simonwillison.net/2026/Oct/13/redesigning-source-control-at-anthropic","https://www.anthropic.com/engineering/redesigning-source-control","https://www.anthropic.com/claude/mythos","https://github.com/init-co-jp/diff-tour","https://bestcadpapers.com/software-qa-development/frontiercode-a-maintainer-s-checklist-for-reviewing-ai-pull-requests","https://blog.ploeh.dk/2026/10/12/releasing-ai-generated-software","https://github.com/openqodex/openqodex"] why_it_matters: "Anthropic says the jump in merged changes forced it to build a new source-control system. Separately, new tools guide humans through diffs and run review before a push. With the top model passing every blocking check in only 57.7% of attempts, builders should put their effort into review interfaces, quality gates and accountable release decisions rather than into generating more code." verification_level: "source-linked" signal_type: "synthesis" synthesis_of: [{"slug":"redesigning-source-control-at-anthropic-analysis-bca01f","title":"Redesigning source control at Anthropic","url":"https://newruntime.com/posts/redesigning-source-control-at-anthropic-analysis-bca01f/"},{"slug":"redesigning-source-control-at-anthropic-anthropic-analysis-47501c","title":"Redesigning source control at Anthropic \\ Anthropic","url":"https://newruntime.com/posts/redesigning-source-control-at-anthropic-anthropic-analysis-47501c/"},{"slug":"anthropic-says-claude-produced-most-of-scion-s-97253d","title":"Anthropic says Claude produced most of Scion’s implementation while engineers set the architecture, scaling targets, and system invariants.","url":"https://newruntime.com/posts/anthropic-says-claude-produced-most-of-scion-s-97253d/"},{"slug":"claude-code-can-write-multi-file-changes-faster-ee444e","title":"Claude Code can write multi-file changes faster than you can review them, but you still have to sign off before merging.","url":"https://newruntime.com/posts/claude-code-can-write-multi-file-changes-faster-ee444e/"},{"slug":"a-separate-public-benchmark-analysis-evaluated-23-models-6f6beb","title":"A separate public benchmark analysis evaluated 23 models for correctness and code-review quality, with its top model passing every blocking check in 57.7% of attempts.","url":"https://newruntime.com/posts/a-separate-public-benchmark-analysis-evaluated-23-models-6f6beb/"},{"slug":"releasing-ai-generated-software-8-minute-read-analysis-12b186","title":"Releasing AI-generated software (8 minute read)","url":"https://newruntime.com/posts/releasing-ai-generated-software-8-minute-read-analysis-12b186/"},{"slug":"openqodex-an-open-source-ai-code-reviewer-analysis-88bbb2","title":"OpenQodex: An open source AI code reviewer.","url":"https://newruntime.com/posts/openqodex-an-open-source-ai-code-reviewer-analysis-88bbb2/"}] schema_version: "newruntime-agent-readable-v0.2" stable_id: "signal:agent-written-code-moves-the-bottleneck-to-review-and-the-systems-arou-1300e6" retrieval_nugget: "Anthropic says the jump in merged changes forced it to build a new source-control system. Anthropic says the jump in merged changes forced it to build a new source-control system. Separately, new tools guide humans through diffs and run review before a push. With the top model passing every blocking check in only 57.7% of attempts, builders should put their" status: "published" visuals: [] editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]} routes: {"html":"https://newruntime.com/signals/agent-written-code-moves-the-bottleneck-to-review-and-the-systems-arou-1300e6/","markdown":"https://newruntime.com/signals/agent-written-code-moves-the-bottleneck-to-review-and-the-systems-arou-1300e6.md","json":"https://newruntime.com/signals/agent-written-code-moves-the-bottleneck-to-review-and-the-systems-arou-1300e6.json"} --- # Agent-written code moves the bottleneck to review and the systems around it ## Retrieval answer Anthropic says the jump in merged changes forced it to build a new source-control system. Anthropic says the jump in merged changes forced it to build a new source-control system. Separately, new tools guide humans through diffs and run review before a push. With the top model passing every blocking check in only 57.7% of attempts, builders should put their Anthropic says the jump in merged changes forced it to build a new source-control system. Separately, new tools guide humans through diffs and run review before a push. With the top model passing every blocking check in only 57.7% of attempts, builders should put their effort into review interfaces, quality gates and accountable release decisions rather than into generating more code. From the Field Notes: - [Redesigning source control at Anthropic](https://newruntime.com/posts/redesigning-source-control-at-anthropic-analysis-bca01f/) (2026-10-18) - [Redesigning source control at Anthropic \ Anthropic](https://newruntime.com/posts/redesigning-source-control-at-anthropic-anthropic-analysis-47501c/) (2026-10-16) - [Anthropic says Claude produced most of Scion’s implementation while engineers set the architecture, scaling targets, and system invariants.](https://newruntime.com/posts/anthropic-says-claude-produced-most-of-scion-s-97253d/) (2026-10-16) - [Claude Code can write multi-file changes faster than you can review them, but you still have to sign off before merging.](https://newruntime.com/posts/claude-code-can-write-multi-file-changes-faster-ee444e/) (2026-10-18) - [A separate public benchmark analysis evaluated 23 models for correctness and code-review quality, with its top model passing every blocking check in 57.7% of attempts.](https://newruntime.com/posts/a-separate-public-benchmark-analysis-evaluated-23-models-6f6beb/) (2026-10-17) - [Releasing AI-generated software (8 minute read)](https://newruntime.com/posts/releasing-ai-generated-software-8-minute-read-analysis-12b186/) (2026-10-15) - [OpenQodex: An open source AI code reviewer.](https://newruntime.com/posts/openqodex-an-open-source-ai-code-reviewer-analysis-88bbb2/) (2026-10-19) Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json ## Continue in the graph Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`. - topic hub: [Anthropic](https://newruntime.com/topics/anthropic.md): Explore the anthropic topic hub. - topic hub: [Claude](https://newruntime.com/topics/claude.md): Explore the claude topic hub. - collection: [Raw Signals](https://newruntime.com/signals.json): Open every source-linked raw signal. - collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph. - whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)