Anthropic's AI Submits 65% of Engineering PRs, Slashes Prompts by 80%
Simon Willison · rss · 2026-07-21
Simon Willison hosted a fireside chat with Anthropic's Claude Code team, revealing how AI coding agents are transforming internal engineering workflows at top AI labs.
Key Insights:
- AI-Driven PRs: The newly launched Slack integration, Claude Tag, now lands 65% of the team's product engineering PRs. The team relies heavily on automated code reviews, manually reviewing only critical changes.
- Prompting Reversal: Stuffing system prompts with examples or negative constraints ("don't do X") degrades performance on newer models. The Claude Code system prompt was recently reduced by 80%.
- Rewrites are Good: Previously a software engineering taboo, rewriting codebases is now highly effective with coding agents if backed by a solid test suite (e.g., internally rewriting Bun in Rust).
- Engineer Role Shift: With build cycles shrinking from months to a week, engineers must pivot from mere execution to cultivating deeper product taste and business sense.
- Agentic Workflows: Claude Code's auto mode is a key enabling technology, and newer models are even competent enough to edit launch videos internally.
More from coding & agent
- Dev builds interactive 3D product experience with GPT-6 Astra + Hyper3D Rodin — nikola_mr64990 · 2026-09-11
- Codex tip: use Sol with Astra and Luna sub-agents to save usage — pvncher · 2026-09-11
- agents-best-practices: a provider-neutral Agent Skill for designing and auditing agentic harnesses — tom_doerr · 2026-09-11
- Cognition's SWE-2 uses a KKT duality argument in RL to shift the effort Pareto curve — YouJiacheng · 2026-09-11
- First-ever Three.js Conference lands in Paris, with a panel on AI-shortened design workflows — OdinLovis · 2026-09-11
- Data engineering, not agent frameworks, is the real bottleneck for enterprise AI agents — dhruv2038 · 2026-09-11