Anthropic launches auto mode in Claude Code to manage permissions
Technology
Anthropic just rolled out "auto mode" for its AI coding assistant, Claude Code.
This new feature lets the AI handle permission decisions by itself, so developers can focus more on coding and less on security pop-ups.
It's meant to make work smoother and safer, moving away from the risky "dangerously skip permissions" option.
Auto mode classifier approves safe commands
Auto mode checks commands using a classifier: safe actions go through automatically, while risky ones get blocked unless you step in.
Right now, it's available as a research preview for Team plan users and will reach Enterprise and API users soon.
It works with Sonnet 4.6 and Opus 4.6 models, but Anthropic still suggests using isolated environments to keep your projects extra secure.