Our team has spent months making cloud sessions (in claude.ai/code or the Cloud option in Claude Desktop's code tab) more reliable. We're incredibly excited for those of you on Pro/Max plans to try this experience by giving you some credits! One of my favorite use cases has been modernizing my board games website that I wrote pre-COVID on my commutes home from work. It's incredible how quickly Opus 5.5 can fix my pre-pandemic bugs and modernized my stack through my phone.

Opus 5.5 is out! This release is notable because the writing quality is so much better. This also translates to artifacts. If you say "Create an artifact to explain X with SVG diagrams + plain terms", you will get amazing infographics.

a few more interesting bits on using projects 🧵 when you first create a project, you give it a goal and a few repositories. the coordinator immediately start getting the lay of the land and will suggest a few threads for you to immediately get started

Super excited to show the world Projects!! We've been building this for the last few months and it has become my new daily driver. A few tips for getting the most out of it 🧵

The new Projects redesign is the tipping point that has converted me from Claude Code CLI to Claude Desktop full time. The threading experience is magnificent. For example, you can ask Claude to fan out work into new threads and track them. You can reference other threads or have Claude locate what you talked about previously. Best of all, you can use this through your Pro/Max plan once Projects are fully rolled out!

New in Claude Code: claude plugin eval See what value your plugin is adding, or if it needs more work. You can create test cases, run your plugin or skill against those test cases, score those runs, then run each case again without the plugin to see the differences.

Check out Alice’s (poteat) new function hooks proposal! It’s a super ergonomic interface and more flexible than the existing set of hook types (including the ability to manipulate the TUI!)

"Ask Maps" in Google Maps is my current favorite non-Claude AI product. Here are some example prompts: > Let's say I'm at ___. Which Target should I go to on my way home? Google Maps will report detour time. You can also ask this *before* you head out. > Seafood restaurant along the coast with a view Led to some great options while I had a short layover in Istanbul recently. It seems that Ask Maps can also combine web results, aggregated (not individual) review data, and realtime traffic/directions data.

It's been so much fun customizing my Claude Tag setup: defining my set of channels, linking my knowledge base repos across those channels (as one access bundle), and setting up routines and artifacts. In fact, I'd say it's been more fun than Claude Code setup from a year ago. The building blocks provide more room for creativity and the end result is so gratifying -- I'm basically building myself a better Slack experience that reduces my cognitive load and gets more done for me in the background.

Rewriting Bun in Rust bun.com/blog/bun-in-rust

Superb article by @lydiahallie that not only covers model vs. effort but also relevant fundamentals of how coding agents work to inform the discussion! It also explains how effort isn’t just for controlling reasoning time. Worth reading even if you think you know coding agents well.

With the launch of Claude Tag, we're sharing the future of work with the world. Working with agents means working in some new forward-looking ways. We're excited to share some best practices below claude.com/blog/building-eff…

Excited to share Claude Tag! Claude Tag is agentic and ambient. It incorporates memory, routines, and emoji reactions to a surface where many of you spend a significant part of your workday. It's a complete evolution from the original Claude integration in Slack.

New in Claude Code: Artifacts. Interactive pages built from your session, like a PR walkthrough or a living project dashboard, shared with your team at a private link. Available in beta on Team and Enterprise plans.

Introducing Claude Fable 5: a Mythos-class model that we’ve made safe for general use. Its capabilities exceed those of any model we’ve ever made generally available.

My colleague's "dynamic workflows" are, in my opinion, the most significant Claude Code innovation in 2026 so far. Unpacking "Claude dynamically writes orchestration scripts" from the blog post: Claude has the flexibility to determine phases, subagents to use, and prompts for subagents, leveraging your session's context *and* Claude's existing ability to write scripts. Workflows massively reduce the prompting you have to do to launch complex agent processing. For reviewing code, I like to set up a workflow with a phase that brainstorms potential issues, a second phase to research each issue in detail, and a third to resolve the issues that are verified through the research. This process gives me much more confidence in my code quality. That said, it will use up lots of tokens.

Amused by this use of hooks as a former Berkeley CS TA 😁

Agent view is an *incredible* experience new experience for TUI. It's super easy to start a new session in any directory you want. The new session uses the directory that the cursor is on. Renaming is extremely critical for maintaining a clean view. Use /rename or set up a UserPromptSubmit hook to rename.

I'm starting office hours for Claude Code's cloud environments. If you use any of our cloud related features on desktop/web/mobile, come talk to me! Bring feature requests, bug reports, or anything in between! Sign up link below.