A Fireside Chat with Cat and Thariq from the Claude Code team

This article summarizes a fireside chat at the AI Engineer World's Fair with Cat Wu and Thariq Shihipar from Anthropic's Claude Code team. Topics include Claude Code, Claude Tag (Slack integration), Fable, coding agent security, evals, tool design, and how Anthropic uses these tools internally. Key highlights: Claude Tag handles 65% of product engineering PRs, Claude Code features are tested internally first, automated code review is increasingly used, and system prompts have been reduced by 80% for newer models. The discussion also covers dogfooding ('ant fooding'), the shift from manual monitoring to delegation, and how agents free up time for creative work.