Everything Anthropic has shipped, in one place.
Long Claude answers now appear more smoothly as they are written, in the browser and the desktop app. Nothing to switch on.
Admins on Claude Team and Enterprise can switch on connected tools for the whole organization through their identity provider, so nobody has to authorize each tool themselves.
Any computer running Claude Code's remote control now shows up in the mobile app, so you can start work on that machine without sitting at it.

Enterprise customers can point Claude at a GitHub repository and get back a list of security flaws with suggested fixes, using Anthropic's security-specialized model.
A setting that makes Claude Code lead with the answer and keep responses short, for Claude Code users.

Four building blocks for agents on the Claude Platform move out of beta, for developers building on the API.
Two changes to Claude Managed Agents: memory in self-hosted sandboxes, and limits on what an agent is allowed to search and read.
Claude Desktop launches about twice as fast as it did a month ago. Small update, most noticeable if you keep it running in the background.

The 50% higher weekly cap on Claude Code stays in place until August 31 for paid plans.

Cowork, the workspace where you hand Claude a task to work through on its own, now runs on phone and web for every paid plan. It was limited to Max users.
