Clauding...

27 changes, but no big stage — version 2.1.206 is a classic stability update. The most interesting detail: /doctor now suggests removing redundant lines from your CLAUDE.md.

Cognition has tested nearly every Claude model inside Devin, its AI software engineer. Claude Fable 5 is the first one the team trusts to run unattended for eight hours — and deliver usable code by morning.

Google DeepMind pushes Gemini 3.5 Pro to July 17 — and rebuilds the model from scratch. The old architecture gets scrapped, a fresh pre-training run started. An expensive signal in the frontier race.

With Muse Spark 1.1, Meta ships its second Superintelligence Labs model — an agentic coder with computer use and a 1M-token context. Zuckerberg broke a three-year X silence to launch it.

With its 'Inviting hard questions' campaign, Anthropic asks the public for its toughest questions about AI — on jobs, families, safety, science. And it promises to publicly track what the company does about them.

Former Fed chair and Nobel laureate Ben Bernanke is the newest member of Anthropic's Long-Term Benefit Trust — the independent body meant to hold the company to its mission. An unusual name for an unusual structure.

As of July 9, 'Reflect' is live in beta — a kind of 'Claude Wrapped' that shows you when and what you use Claude for. It comes with quiet hours, break nudges, and one question that sticks: what do you want to keep doing yourself, even if Claude could do it faster?

The July update to VS Code brings a small but practical change for Claude users: multiple parallel chats within a single agent-host session. Sounds unremarkable — but it changes how you work with the agent.