**[Shopify CEO Pressures Anthropic: May Consider Internal Ban if Claude Code Refuses to Support AGENTS.md]** Beating AI News Flash: Shopify CEO Tobi Lütke publicly stated that if Claude Code continues to refuse to read AGENTS.md and .agents/skills, he is considering banning Claude Code internally within Shopify. Tobi pointed out that as AI programming tools like Codex, Cursor, and Claude Code are simultaneously used by corporate teams, an increasing number of coding agents are beginning to support unified AGENTS.md configuration standards, testing processes, and agent directives. However, Claude Code currently relies primarily on CLAUDE.md and .claude/skills, which may lead to different agents interpreting the same code repository with varying rules, creating what is referred to as "split brain." He believes that while maintaining multiple configuration files has limited impact on individual developers, for enterprises like Shopify with a large number of engineers, multiple agent contexts require continuous synchronization. Any discrepancies could result in agents deviating from team standards during execution. This controversy appears to be a dispute over configuration file formats on the surface, but it fundamentally involves the question of whether AI coding agents in enterprises should adopt vendor-specific standards or establish cross-agent universal standards for project specifications, skills, and context. [Original Link]