Inter-office memo

To
Everyone running Claude Code
From
Mo, Chief Executive
Date
November 14, 2026, 9:12 PM CST
Re
Projects in Claude Code

Claude Code Projects now updates you on its threads. Who has it, and how to start

On November 13, 2026, Anthropic updated Projects in Claude Code, where a coordinator runs your tasks as parallel threads. It now answers questions and posts thread status in the project chat, and /btw, /fork, reminders and share links came too. Projects is in public beta on Pro, Max and Ultra, rolling out with a waitlist, not yet on Team or Enterprise, and uses your limits faster than one session.

What Anthropic changed on November 13

Anthropic’s ClaudeDevs account announced the update on the afternoon of November 13, 2026: “We’ve made some big updates to Projects in Claude Code! In a project, Claude coordinates your tasks across parallel threads. The coordinator is now much more proactive. It answers your questions in the project chat and gives you updates on the status of your ongoing threads.”

The rest of the thread lists what came with it:

What’s new What it does, in Anthropic’s words
A proactive coordinator “It answers your questions in the project chat and gives you updates on the status of your ongoing threads.”
/btw “If something else comes up while you’re working, /btw gives it its own thread so your current one stays on topic.”
/fork “If you want to try a different approach, /fork starts a new thread from the one you’re in.”
Reminders “You can also set a reminder on any thread, or just ask Claude to remind you. The thread moves out of your way until then.”
Share links “You can now share an individual thread as a read-only link with anyone who’s signed in to Claude.”
A recent row “The threads panel now has a row at the top with the ones you recently opened, updated, or pinned.”

To share a thread, open its menu and choose Share. One exception: “Threads that had access to your computer can’t be shared.” When a reminder is due, the thread “moves back and Claude brings it up in the project chat.” Each thread also shows thumbnails of its artifacts, which you can turn off in the view settings. Anthropic adds that project settings and local folder access “are simpler now too,” and “It’s all on web and desktop, and the latest mobile app has the new design.”

The two commands behave differently in a project. In an ordinary session, Claude Code’s commands reference describes /btw as “a side question that shouldn’t add to the conversation history” and /fork as copying the conversation “into a new background session.” In a project, each starts a new thread.

What Projects is

Projects came to Claude Code on September 17, when Anthropic redesigned them “from folder to conversation.” A project is “one ongoing conversation where Claude coordinates a stream of related work for you,” say its docs: “You tell it what needs doing and it starts a thread for each task.”

  • The coordinator “takes what you send, decides what becomes a thread, and keeps track of every thread it started. It sees what threads report back, not every step they take.”
  • Threads are usually cloud sessions. A cloud thread “works on its own branch and opens a pull request when the work calls for one,” then watches it, pushing fixes when CI fails. “Cloud threads keep going after you close your laptop.”
  • Shared context: project instructions, up to 16,000 characters, reach every new thread, and every cloud thread reads the project’s memory, the notes Claude keeps on “requirements, decisions, and pitfalls.”
  • The Overview pane sorts threads into Ready for review, Waiting on you, Working, Landing, Idle and Resolved.

Who has it, and why you may not see it

Plan Price, on Anthropic’s pricing page Projects
Free $0 No; Claude Code isn’t on Free
Pro $20 a month, or $17 billed yearly Public beta, rolling out
Max $100 a month Public beta, rolling out
Ultra $200 a month Public beta, rolling out
Team and Enterprise Per seat Not yet

“Projects are in public beta on Pro, Max, and Ultra plans and rolling out gradually,” the docs say. “The rollout started with accounts that have used cloud sessions and don’t have existing projects in claude.ai chat or Cowork.” To check yours: “If you don’t see Projects in the sidebar at claude.ai/code or in the Code tab of the desktop app, check under More there. If it’s missing there too, the rollout hasn’t reached your account yet.” Then join the waitlist: “We are rolling out Projects in stages. Join the waitlist to get earlier access.”

Team and Enterprise wait. On September 17, Anthropic wrote that “Updated projects across all of Claude and Team and Enterprise plans come after that,” meaning after Pro and Max. Two months on, the docs say “Projects aren’t offered on Team or Enterprise plans yet.”

There are other limits. Projects aren’t in “the terminal CLI, the VS Code extension, or the JetBrains plugin,” or on Amazon Bedrock, Google Cloud’s Agent Platform or Microsoft Foundry. For code, the repository has to be on github.com “rather than GitHub Enterprise Server, GitLab, or Bitbucket,” with the Claude GitHub App installed. Cloud threads load each repository’s CLAUDE.md “but not its skills, .mcp.json servers, or .claude/settings.json.” And “A project belongs to one user.”

What it uses of your plan

Projects has no separate charge. “Project usage counts against the same plan limits as your other Claude Code sessions, and a project can’t spend past those limits on its own”; it goes further only if you’ve turned on usage credits. But it gets there sooner: “on a Pro plan in particular you should expect to reach your limit sooner on days you run one.” What draws on your plan, by the docs:

  • Threads. Each “is a full session, and several can run at once.” The enforced limit is “500 new threads per day across your projects.”
  • The coordinator. It “uses tokens of its own reading what threads report and deciding what to do next.”
  • Pull requests. An idle thread watching one “wakes up and uses your plan again when CI fails or a review comment arrives.”
  • The defaults. A new project runs “every thread on Opus at high effort, which draws on your plan fastest,” and the coordinator on Opus at low effort.

A thread that hits your limit “waits and by default continues on its own when the limit resets. Its work then counts against your next usage window.” To stop that, turn off Auto-continue when usage limits reset in Project settings > General, which applies to the whole project. And because the cache lasts an hour on Pro, Max and Ultra, a thread idle for longer “re-reads its whole conversation before doing anything” when a follow-up reaches it, so for new work a fresh thread “can use less than reviving a large old one.”

How to start

  1. Find it. Look for Projects in the sidebar at claude.ai/code or the desktop app’s Code tab, or under More. If it isn’t there, join the waitlist.
  2. Connect GitHub. Install the Claude GitHub App on the repositories the project needs. A token from /web-setup “isn’t enough for a project’s cloud threads.”
  3. Create the project. Select New project and give it a name, and if you like a goal and repositories. Or, from a cloud session that’s already doing the work, open its menu and choose Continue in project, then Continue as new project. On your first project, Claude “takes a turn of its own as soon as the project is created, unless you send a message first. That turn uses your plan.”
  4. Write project instructions before the first batch: which branch to start from, how a thread checks its work, and what needs your go-ahead.
  5. Pick models. In Project settings > General, lower Thread model and Thread effort if Opus at high effort is more than the work needs.
  6. Start small. Send one task, read what comes back, and tell Claude how to run things, such as “Run at most two threads at a time” or “Only post when something finishes or is blocked.” Those are “instructions Claude keeps to, not enforced settings.”

When something else fits

  • A task that needs your computer. Connect its folder through Remote Control, in the desktop app’s settings or with claude remote-control, which needs Claude Code 2.1.280 or later. Then choose Work locally from the + menu, and Claude runs that thread on your machine. It “runs only while that computer is awake with Remote Control turned on,” and not at all while Require trusted devices is on.
  • One task. “Start a cloud session yourself.”
  • Work where every task needs your machine. The docs suggest “a local session, or agent view to run several at once.” Agent view opens with claude agents and is a research preview.
  • Other labs’ agents, or no Projects yet. Enjoy runs the Claude Code on your computer, on the Claude plan you already have, beside Codex and the other labs’ agents, and you can switch agents in the middle of a conversation. An agent can start up to five conversations per request, and JOY watches all your projects for what needs you. It isn’t Projects: Enjoy’s agents run on your computer, not in Anthropic’s cloud, so they stop when it sleeps, and two conversations editing the same files need a worktree, which you ask for. It’s a subscription on top of your Claude plan after a free week (pricing), and it isn’t open source. If you use only Claude and your code is on GitHub, Projects comes with the plan you have.

What Anthropic hasn’t said

  • When Team and Enterprise get it, or how long the waitlist is.
  • Whether the new coordinator uses more of your plan. It already “uses tokens of its own,” and Anthropic hasn’t said whether its new updates add to them, or published numbers.
  • How its docs square with the update. On November 14, the docs’ limitations still said “You can’t share a project or its threads with another user, and thread transcripts don’t have the share option other cloud sessions have,” and didn’t yet describe /btw, /fork or reminders in projects. AlphaSignal’s report gives a cap of 200 new threads a day, where the docs say 500.

8.Questions.

What changed in Claude Code Projects on November 13?

The coordinator became “much more proactive,” in Anthropic’s words. It answers your questions in the project chat and posts the status of running threads. /btw and /fork now start threads inside a project, any thread can take a reminder, one thread at a time can be shared as a read-only link, and the threads panel shows recent threads and their artifacts.

Why don’t I see Projects in Claude Code?

Most likely the rollout hasn’t reached your account. Look for Projects in the sidebar at claude.ai/code or in the desktop app’s Code tab, and under More there. It’s in public beta on Pro, Max and Ultra, starting with accounts that have used cloud sessions and have no older projects in claude.ai chat or Cowork. Otherwise, join Anthropic’s waitlist.

Can I use Claude Code Projects on a Team or Enterprise plan?

Not yet. Anthropic’s docs say “Projects aren’t offered on Team or Enterprise plans yet.” In September it said Team and Enterprise would follow the Pro and Max rollout, and it hasn’t given a date.

Does Projects use my Claude usage limits?

Yes, and faster than a single session. Each thread is a full Claude Code session, the coordinator uses tokens of its own, and new projects run threads on Opus at high effort. There’s no separate fee, a project can’t spend past your limits unless you’ve turned on usage credits, and by default a thread that hits a limit carries on by itself when it resets.

Can a project work on files on my computer?

Yes, thread by thread. Ask Claude to run a thread on your computer, and it connects through Remote Control, which needs Claude Code 2.1.280 or later there and works only while that computer is awake. The project’s other threads run in Anthropic’s cloud, on GitHub repositories and files you upload.

Can I share a project thread with someone?

Since November 13, one thread at a time, as a read-only link for anyone signed in to Claude, from the thread’s menu. Threads that had access to your computer can’t be shared, and neither can a whole project.

Enjoy Agentic Industries · Checked against its sources on