Anthropic Rebuilds Claude Code Around Fleets of Cloud Agents
Anthropic's revamped Claude Code Projects lets one developer command a fleet of cloud agents. For India's vast coding base, that reshapes the billing math.
The News
Anthropic has rebuilt the Projects feature inside Claude Code, letting developers run several AI coding agents at once inside a single shared workspace. The revamped tool went into beta on 17 September for a limited group of Claude Pro and Max subscribers, with wider access promised soon after.
Each project now carries a shared memory, a set of goals, and a common library of files and artifacts. Work is divided into "threads" that run different tasks in parallel, while a "coordinator" directs the overall effort. Under the hood, every thread is a Claude Code cloud session operating on its own branch and its own copy of the repository.
When two threads touch the same code, Anthropic resolves the overlap as a merge conflict, exactly as any pull request would. Individual threads can split their assignments into smaller pieces using subagents, loops, and workflows so that heavier jobs finish faster. Developers can either message each thread on its own or steer the whole fleet from one project chat.
At launch the threads run entirely in the cloud, though Anthropic says support for local tools and code is arriving "very soon." The company plans to widen access to all Pro, Max, Team, and Enterprise users, and eventually to Cowork and ordinary Claude chats.
Why It Matters
The release marks another step away from single-assistant coding help towards orchestration, where one person supervises a group of semi-autonomous workers rather than typing alongside a single bot. Anthropic itself compares the tool to Grok Bot and others built to manage clusters of agents, a sign that the battleground is shifting from raw model quality to how well a system coordinates many models at once.
The last time developer tooling moved this sharply was the arrival of GitHub Copilot in preview in June 2021, which pulled machine-written suggestions directly into the editor. Projects pushes further: instead of finishing a line, the software parcels out an entire backlog, tracks branches, and reconciles clashes without a human holding every thread. Running fleets of cloud sessions also burns far more compute than a single chat window, and that cost will decide whether orchestration stays a premium feature or becomes standard.
Indian Angle
India has an outsized stake in this shift. GitHub has projected that the country will host the world's largest developer community before the end of the decade, overtaking the United States, which means any change to how code gets written lands on a bigger base here than almost anywhere else.
The immediate pressure falls on the services industry. Firms such as TCS, Infosys, Wipro, and HCLTech, along with the global capability centres in Bengaluru, Hyderabad, and Pune, sell billable engineering hours. A tool that lets one engineer supervise a fleet of cloud agents compresses the very metric those contracts are priced on, hastening the move towards outcome-based billing.
There is a currency catch too. Cloud sessions are metered and charged in dollars, so a rupee-denominated startup in Bengaluru pays the full foreign-exchange premium on every parallel thread it spins up. That math will keep Indian founders watching domestic and open-weight alternatives closely, even as the productivity case for orchestration grows harder to ignore.
FAQ
When does Claude Code Projects become available?
The revamped Projects feature entered beta on 17 September 2026 for a selected set of Claude Pro and Max subscribers. Anthropic says it will expand access to all Pro, Max, Team, and Enterprise users, and later to Cowork and standard Claude chats, though it has not given firm dates for each tier.
How is this different from a normal coding assistant?
A standard assistant helps one developer in real time. Projects instead runs multiple "threads," each a separate cloud session on its own branch, under a "coordinator" that assigns and reconciles the work. It behaves more like a team lead directing several workers than a single autocomplete tool.
Can Indian teams use it on local code?
Not yet. At launch every thread runs in the cloud. Anthropic says support for local tools and code is coming "very soon," which matters for Indian enterprises and banks bound by data-residency rules that limit what source code can leave their own systems.
Where can I read the original announcement?
The launch was reported by The Verge, whose coverage details how threads, coordinators, and merge-conflict handling work. The full write-up is linked in the attribution below.
This story was reported by The Verge. Read the full original coverage at The Verge.