Editorial illustration for Claude Code Projects Coordinate Multiple AI Agents in the Cloud
Claude Code Projects Now Coordinate Multiple AI Agents
Anthropic is giving Claude Code a rewrite of its Projects feature, and the pitch is less "chatbot" and more "engineering team." Starting today, select Claude Pro and Max subscribers can run a beta version that lets one project juggle multiple AI agents at once, each working from a shared pool of memory, goals, and files. Anthropic plans to roll this out to all Pro, Max, Team, and Enterprise accounts later, along with Cowork and standard Claude chats.
The setup mirrors what's already happening elsewhere in the industry: tools like Grok Bot have pushed the same idea of managing a roster of agents instead of one assistant at a time. In Claude Code's version, each project spins up "threads" that tackle separate tasks in parallel, while a coordinator agent oversees the whole operation and keeps everyone pointed at the same objective. That structure raises a practical question Anthropic has had to answer directly: what happens when two agents, working independently, end up touching the same piece of code at the same time.
The revamped projects feature in Claude Code allows users to run multiple agents under the same roof, with a shared memory, goals, and library of files and artifacts.
Why this matters
Anthropic is betting that the bottleneck in AI-assisted coding isn't model quality anymore, it's coordination. Giving each thread its own branch and repo copy is a sane way to let agents work in parallel without one overwriting another's progress, but that "coordinator" is doing a lot of heavy lifting the summary doesn't fully explain. What happens when two threads touch the same file?
The quote trails off right at the interesting part, and that's the detail engineering teams actually need before they hand real production repos to a swarm of cloud agents. For founders building on Claude Code, this is a signal that Anthropic wants to own the orchestration layer, not just the model, putting it in more direct competition with Grok Bot and whatever OpenAI ships next in this space. Worth testing on a messy, multi-file refactor before trusting it on anything customer-facing.
The pitch is compelling. Whether the merge conflicts get resolved cleanly, or just quietly deferred to a human at 2 a.m., is the part we'd want to see in practice.
Related Reading
Browse all Anthropic news →
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
Related Topic Claude gains shared context in Excel, PowerPoint; Microsoft adds Copilot Cowork
AI Tools & Apps
Related Topic Anthropic: Hackers Stole Claude Tokens Via Compromised Session Key
LLMs & Generative AI
From Archives Base Labs Launches Open-Weight AI Safety Partnership
From Archives AI Agents Raise Ownership Questions at TechCrunch Disrupt 2026
From Archives Anthropic and OpenAI Propose Safety Evaluator Access to Staff
Research & Benchmarks
From Archives Anthropic Merges Claude Chat and Cowork Into Single Product
LLMs & Generative AI
Common Questions Answered
How does Claude Code's revamped Projects feature allow multiple AI agents to work together?
The revamped Projects feature enables multiple AI agents to operate simultaneously under the same project, sharing a common pool of memory, goals, and files. Each agent works from its own branch and repository copy to prevent agents from overwriting each other's progress while maintaining coordination through a shared workspace.
Which Claude subscribers can currently access the beta version of the new Projects feature?
Currently, select Claude Pro and Max subscribers have access to the beta version of the revamped Projects feature. Anthropic plans to expand availability to all Pro, Max, Team, and Enterprise account holders, as well as users of Cowork and standard Claude chats at a later date.
What does Anthropic believe is the main bottleneck in AI-assisted coding that Projects addresses?
Anthropic believes that coordination between AI agents, rather than model quality itself, is the primary bottleneck in AI-assisted coding. The revamped Projects feature addresses this by enabling multiple agents to work in parallel with proper coordination mechanisms, allowing engineering teams to leverage multiple AI agents more effectively.
How does the Projects feature prevent conflicts when multiple agents work on the same files?
The Projects feature gives each agent its own branch and repository copy, which allows them to work in parallel without one agent overwriting another's progress. This architectural approach enables safe concurrent work while maintaining a shared memory and goals structure for coordination.
Further Reading
- Claude Code's revised projects adds AI orchestration, but local developers must wait - ZDNet
- Anthropic updates Projects in Claude Code for AI agent coordination - UA.News / The Verge coverage
- Claude Code by Anthropic | AI Coding Agent, Terminal, IDE - Anthropic
- Run agents in parallel - Claude Code Docs - Anthropic Docs
- Your Home for Multi-Agent Development - Visual Studio Code
