How to use Claude Code Projects (open to Pro and Max)

What changed on October 9

Claude Code Projects went into beta on September 17, 2026 for a small group of Pro and Max subscribers. On October 9 the official Claude Developers account posted that every Pro and Max user on the waitlist had been let in. The feature is still labelled beta, and Anthropic's documentation says it is not available on Team or Enterprise plans yet.

If you are on Pro or Max and do not see Projects in the left sidebar at claude.ai/code or in the Code tab of the desktop app, the rollout has not reached your account. You can join the waitlist at claude.com/form/projects.

What a project is

A project is one long-running conversation where Claude coordinates a stream of related work. You type what needs doing, and Claude starts a thread for each task. A thread is a separate Claude Code session with its own context window. Threads run in parallel in the cloud, so five tasks become five threads working at the same time.

• Cloud threads keep going after you close your laptop. You can check on them and steer them from the Claude mobile app.
• Every thread starts with the same context: the project's repositories and files, its instructions, and its memory.
• The Overview pane next to the conversation shows which threads are ready for review, which are waiting on you, and which are still working.

Check you have what you need

• A Pro or Max plan, with Projects showing in your sidebar at claude.ai/code, in the desktop app's Code tab, or in the Claude app for iOS and Android.
• Only if the project will work on code: the code must be on github.com (not GitHub Enterprise Server, GitLab or Bitbucket), your connected GitHub account needs push access, and the Claude GitHub App must be installed on each repository. Install it from github.com/apps/claude.
• No code? Skip GitHub. A project can run on uploaded files alone.

Create your first project

1. Go to claude.ai/code (or open the Code tab of the desktop app) and select Projects in the left sidebar, then New project.
2. Name it after one stream of work you will keep adding to, for example "Client contracts" or "Payments API".
3. Add a one-line Goal (optional), such as "Hold p95 API latency under 200 ms" or "Find every renewal date and payment term in these contracts".
4. Under Context, click Add for each GitHub repository, file, folder or Google Drive folder the threads should read.
5. Click Create project.

On your first project Claude takes a turn of its own as soon as it is created, unless you send a message first. It may start one read-only thread to explore a repository and post setup recommendations. That turn uses your plan.

Write your project instructions once

Project instructions are the brief every new thread starts from, up to 16,000 characters. Click the gear icon in the project header, then Memory > Project instructions. Anthropic's documentation suggests covering what the project is for, where the work happens, how a thread checks its own work, what to do when something is missing, and what needs your go-ahead first. A short version for a non-code project:

This project reviews client contracts for Harbor & Pine Studio.
- Deliver each write-up as a file on the Library tab.
- Quote the exact clause and page for every finding.
- If a file is missing or unreadable, say so in your first message and stop.
- Never email anyone or change a document without asking me.

When you correct a thread later, also tell Claude to remember the correction. It goes into project memory and later cloud threads start with it.

Send work and read the results

Give Claude tasks in the project conversation, one at a time or several in one message. A quick question usually gets an answer in the conversation. New work goes to a new thread, or to a thread already working in that area, and Claude tells you which. Each thread shows up as a card under your message.

The Overview pane groups threads by state: Ready for review, Waiting on you, Working, Landing, Idle and Resolved. Files a thread produced are on the Library tab. In the desktop app you also get a notification when a thread needs your input or hits an error.

You can tell Claude how to run the project in plain words, for example "Propose threads and wait for my go-ahead before starting them", "Run at most two threads at a time" or "Only post when something finishes or is blocked". Claude saves these preferences to project memory.

Use it without any code

Anthropic's own examples include a folder of contracts and a support-ticket export you keep coming back to with new questions, such as "find the ten most common integration mistakes in these tickets". Upload the documents instead of adding a repository, and threads deliver each write-up as a file on the Library tab.

• From the Library tab you can add up to 100 files and 2 GB in one pick, with a single file up to 480 MB.
• Files over 30 MB are skipped in the New project dialog, so add large files from the Library tab after the project exists.
• Uploads are copies. If you change a file on your computer, upload it again and choose Replace.

Keep your plan usage down

A project draws on the same plan limits as your other Claude Code sessions and uses them faster, because each thread is a full session. A new project runs every thread on Opus at high effort by default. Three settings to check in Project settings > General before you send a big batch:

• Thread model and Thread effort: pick a smaller model or lower effort for routine tasks. You can still ask for a bigger model in one task.
• Coordinator model and Coordinator effort: these apply to Claude in the project conversation itself.
• Usage: the Usage section shows token use by thread and by model.

Start with one small task, open the thread when it finishes to see how it reports back, and only then send the batch.

Limits to know

• Cloud threads do not see anything on your own computer. For a task that needs a local file, database or tool, you can ask Claude to run that thread on your computer through Remote Control (Claude Code v2.1.280 or later, with the desktop app open or claude remote-control running).
• A project belongs to one user and cannot be shared with other users.
• Team and Enterprise plans do not have the new Projects yet.
• It is a beta. Anthropic says threads run in auto mode when the model supports it, so most tool calls run without asking you. Put anything that needs your go-ahead in the project instructions.

Questions people also ask

Sources

• Claude Developers on X, October 9, 2026: every Pro and Max user on the waitlist let in
• Let Claude coordinate ongoing work with Projects (Claude Code documentation)
• Projects redesigned: from folder to conversation (Anthropic, September 17, 2026)
• Claude Code Projects waitlist
• Claude GitHub App
• Remote Control (Claude Code documentation)