Articles from claude.dev
5 kept
- Using a "chaos loop" to harden the Claude Docs launch: 60 server errors across billions of requests / claude.dev Blog
Anthropic used a two-week autonomous 'chaos loop' driven by Claude to harden the Claude Docs service, surviving a launch of 4.6 million docs with only 60 server errors across 2 billion requests.
- How the Claude Code team thinks about verification / claude.dev Blog
The Claude Code team describes using E2E, property, and differential tests to verify AI-generated code, along with processes for keeping verifiers reliable and cost-effective.
- How the Claude Code team does code review / claude.dev Blog
The Claude Code team uses layered automated reviews, AI-driven checks, and risk-based human approvals to handle massive code volume while keeping revert rates low.
- Automating eval design and hillclimbing with Claude / claude.dev Blog
The article explains how to design reliable evaluations and hillclimb model performance using Claude Code’s claude-api skill, showing principles and examples of cost and accuracy gains.
- How we made claude.ai 3x faster in two weeks / claude.dev
Claude and engineers made claude.ai 3x faster in two weeks by using Claude to build benchmarks, find bottlenecks, and ship 3,000 changes safely.