Did you know that we have LinkedIn, Instagram and X accounts
that you can follow?

In this issue:

🤿 DEEP DIVE

Claude Code's New Boss Is Also Claude

The threads do the coding. One Claude keeps track of everything.

Coding agents doing several things at once isn't really the unusual part anymore.

The interesting bit here is that Anthropic has put another Claude above them.

Okay, so Anthropic just launched Claude Code Projects, and the easiest way to understand it is probably as a coordinator for a development job that doesn't neatly end when one coding session does.

You give it the larger thing you're working on. Claude breaks that into separate threads, those threads go off and do the work, and the coordinator keeps track of the broader goal while you continue talking to it.

Anthropic's own description is pretty good.

"Threads do the work, Claude directs it."

That's actually the important bit

Say you're trying to reduce an app's checkout latency.

Rather than manually opening separate Claude Code sessions to profile endpoints, test optimizations and work on pull requests, the Project can create threads for those jobs and coordinate them under the same larger objective.

And you don't have to wait until one finishes before giving it something else.

You can keep briefing the coordinator as things change. It decides whether the new instruction belongs in an existing thread or needs a new one. If you want to see what's happening underneath, you can open an individual thread and steer it directly.

It's less like having one AI programmer and more like having a conversation with the thing organizing several of them.

The memory might matter more than the parallelism

This is where Claude Code Projects gets more interesting.

All those worker threads contribute to and draw from shared Project memory. So if a release date moves, a feature gets dropped, or there's an instruction about who needs to be consulted before touching a service, that decision can stick around as the work continues.

That's a pretty different problem from simply getting an agent to write some code.

It's about not having to reconstruct the state of a long-running project every time you come back to it.

And yes, Anthropic already has something called Claude Projects, which makes the naming slightly confusing.

The distinction in the article is basically this. Claude Projects keeps chats, files, instructions and knowledge together. Claude Code Projects is designed to keep the work coordinated across multiple coding sessions.

For now, those workers live in the cloud

There is an important limitation.

The worker threads currently run as cloud sessions, so they can't access files or tools on the developer's computer or systems on an internal network. Anthropic says local support is coming "very soon."

The beta is also limited initially to select Claude Pro and Max subscribers using Claude Code cloud sessions, with access expected to expand over the following week.

So this isn't just "Claude Code, but more agents."

The bigger idea is one continuing conversation sitting above the agents, remembering what the project is trying to accomplish while deciding where the next piece of work should go.

The quick take

🤝 FROM OUR FRIENDS

A quick share we thought you might find useful

Two free days to see exactly how AI, automation, and modern tools are changing the way people build and grow online, with real sessions, real builders, and real takeaways you can use right away.

The way people build and work online has changed fast. AI tools are compressing what used to take days into hours, and the builders keeping up aren't just using AI, they're rethinking their whole workflow around it.

That's what Prepathon 2026 is about. It's a free virtual event from Cloudways by DigitalOcean, running September 22–23, packed with hands-on sessions on AI-assisted building, automation, and what's actually working right now for developers, agencies, and growing businesses.

Expect practical sessions, live activities, and real examples you can apply, not abstract trend talk.

It's also happening in person this year in Karachi, Mexico City, and Bengaluru, for anyone who wants it live.

See you there.

🔥 WHAT’S THE BUZZ

What people are building, sharing, and talking about

One thing that can help with longer AI projects is having one agent decide what to do next and another agent actually do the work. It’s a simple change, but it keeps things more organized and helps the AI stay on track.

⚒ TOOL SNAPSHOTS

Futuristic tools within AI, no-code, and productivity

👀 ON OUR RADAR

One more thing we thought you'd like

Ava is the first AI BDR to run outbound end to end, from finding leads to booking meetings, autonomously or on copilot. She's SOC 2 audited and trusted by companies like DoorDash and Grammarly. Book a demo.

🤖 AI MADE THIS

Interesting and inspiring creations made with AI

AI as Artist

A gallery installation presents three fictional AI-born artists with their own interviews, publications, exhibition histories, and work spanning sculpture, images, video, and code. It’s basically asking what happens when AI stops being framed as a tool and gets given an artistic identity of its own; the project was also exhibited at Ars Electronica 2026 as part of Art After AI.

AI used
The page confirms that the project includes AI-born artists working across image, video, code, and 3D printed sculpture, but it does not provide a complete model-by-model production breakdown.

ℹ️ ABOUT US

The Intelligent Worker helps you to be more productive at work with AI, automation, no-code, and other technologies.

We like real, practical, and tangible use-cases and hate hand-wavy, theoretical, and abstract concepts that don’t drive real-world outcomes.

Our mission is to empower individuals, boost their productivity, and future-proof their careers.

We read all your comments - please provide your feedback!