(2026-01-13) Zvim Claude Coworks

Zvi Mowshowitz on Claude Coworks. Claude Code does a lot more than code, but the name and command line scare people. Anthropic realized a rebrand was in order. Two weeks later, we have Claude Cowork, written entirely by Claude Code.

This is still very much a research preview, available only for Claude Max users on Macs with a bunch of bugs and missing features. It will improve rapidly over time.

Cowork combines a lot of the power of Claude Code with the ordinary chat interface, giving it access to a folder on your computer and to Claude Code’s planning and agentic capabilities. It can use that folder as context, to download, to organize and create files, and it can be paired with Claude for Chrome and use your existing connectors.

The system prompt is here, the core non-tooling parts seem unchanged. This post will cover Claude Cowork, and also updates since last week on Claude Code.

Early Takes

What exactly can it do at this early stage?

Claire Vo: It’s basically local Claude Code with a Mac OS app wrapper focused on a few core primitives:

• Connectors / MCPs - external services Cowork has access to
• Filesystem - runs locally so will create/read things on your file system
• TODOs/Steps - discrete trackable steps cowork will take to execute your tasks
• Artifacts - files generated in the process of doing your task
• Context - files / sources / connectors used when doing your task
• Skills - preloaded with a few key skills, esp. file type creation ones like DOCX, PPT, etc. Claude generally has access to these, so not new.

Minimum Viable Product

It’s early days, and she reports there were still some other kinks being worked out. In particular, the connectors are having problems.
Tibor Blaho: Available now for Max subscribers on macOS desktop app only, with no project support, no memory between sessions, no sharing, app must stay open during tasks, and consumes more usage than regular chat, with plans to add cross-device sync and Windows support

Maximum Viable Product

There is a huge overhang in AI capabilities.

Thus, a common pattern is that someone figures out a way to do useful things at all that humans are willing to learn how to use. And then we muddle down that road, and it’s not first best but it still wins big.

That’s what Claude Code was, and now that’s what Claude Cowork will be for normies. Presumably OpenAI and Google, and then others, will soon follow suit.

Backup Your Directory

If you’re worried Claude Cowork or Claude Code will delete a bunch of stuff in a directory, and you don’t want to use a full virtual sandbox solution, there’s a rather simple solution that also works, which is: Backup the directory, to a place Claude can’t get to it. Then if the worst happens, you restore the backup.

Meanwhile In Claude Code News

The latest guide to Claude Code, feedback seems very good. Key highlights:

• Think before you type. Enter plan mode first and go back and forth a lot.
• Keep claude.md short, max 50-100 instructions. Use # while working to edit it.
• Store things in external files.
• Try to only use 30% of the context window, after that performance degrades.
• Make your prompts as specific as possible, including what not to do.
• Try out various hooks, MCPs, you name it. Experiment.
• When stuck, be creative, pivot, simplify, clear the conversation and start again.
• Build systems, not one-off tasks.

Nikhil Krishnan: I’ve spent the last 48 hours in Claude Code - as a non-technical person it’s basically unlocked three very big things for me.

• The ability to interact with APIs generally - again, as a non-technical person one of the big barriers to running the business has been touching APIs. For example, what you can do in Stripe in the non-developer portal vs. through the API is night and day.
• The ability to thread things together - another issue has been threading several different products we work with together to do cohesive tasks. Zapier gets you part of the way for triggers, but Claude Code let’s me do way more complex things that touches multiple things simultaneously
• Run something regularly - being able to set a script and run it regularly with this level of ease is a game changer. In about an hour I set up a daily email to myself that tells me the top 3 emails I need to respond to based on a priority scoring system we made together that pulls data from a few different places.

I’m not there yet, largely because we think in different ways, but largely because I’m just getting started with ‘oh right coding things just happens, do coding agent shaped things.’

The coding agents will still produce the most value for professional coders, because they can go into supercharged mode with them and get the most out of them, but that requires the professionals to swim upstream in ways the rest of us don’t have to.

As Alex Albert puts it, you get to stop thinking doing something is ‘not worth your time,’ or for Simon Willison entire features are no longer ‘not worth your time’ at least not until they run into serious trouble.

You Are The Bottleneck

She and I are very similar types of programmers: Kelsey Piper: In college, I was once told that the really hard part of programming was knowing, in sufficient detail, what you wanted the computer to do. This was not my experience of programming.
In my experience of programming, the really hard part was figuring out which packages weren’t installed or weren’t updated or were in the wrong folder, causing the test we’d done in class to completely fail to work in the same way on my own computer. The next really hard part was Googling everything the debugger spat out to find an explanation of how to make it go away.
… Claude Code solves all of that. Programming, now, really is just a matter of knowing in sufficient detail what you want the computer to do.
… Now, 99% of the time, it feels like magic. The remaining 1% is absolutely maddening.

Claude Code Upgrades

Anthony Morris ツ: We shipped A LOT of updates to Claude Code on desktop in the last week.

Numman Ali says v2.1.3 has ‘solved the compaction issue’ so long as you use planning mode and explicitly ask the model for a comprehensive TODO list. It’s hard to tell, but I’ve certainly blown over the compaction line on many tasks and when I’ve saved the necessary context elsewhere it’s mostly turned out fine.

Oh No

In the interest of not silencing critics, Holly Elmore claims I’m bad now because I’m enthusiastic about getting use out of Claude Code, a ‘recursively self-improving agent.’
I affirm David Manheim’s response that there is no reason for an individual not to use such tools for their own purposes, or not to get excited about what it can do outside of potentially dangerous forms of self-improvement.

I’m In Danger

There is a new method called the ‘Ralph Wiggum’ (Ralph loop) technique, where you tell Claude Code continuously to ‘improve the code’ it has already written. Some say it works great, but the name does not inspire confidence.

The world is collectively underinvesting in optimizing and standardizing such techniques.

In The Beginning Was The Command Line

What is the difference between a command line and a chat interface?

The main real difference is that the terminal makes it annoying to edit prompts?
It’s almost entirely about perception. One feels like talk with an entity, one like commands and bash scripts. One looks like a slick modern UI, the other a stark black text box

There is also a clear plan to have different system prompts, and to build in a different more user friendly set of default connectors and tools.

That plus the change in perception could be a really, really big deal.