Development & Automation

Even When You Hit the Usage Limit, Claude Code Doesn't Stop.

The Claude Code desktop app has a new auto-continue checkbox. When your limit resets, the work you were doing picks up on its own.

· Huuush

Even When You Hit the Usage Limit, Claude Code Doesn't Stop

Key takeaways

  1. 1When you use Claude Code, you hit the usage limit. At that moment, work stops.
  2. 2You had to check for yourself when the limit would lift. Then you had to restart manually.
  3. 3That flow has changed recently. @ClaudeDevs is Anthropic's official account. The account announced the new feature.
A photo of a brass stopwatch next to a toggle switch

When you use Claude Code, you hit the usage limit. At that moment, work stops.

You had to check for yourself when the limit would lift. Then you had to restart manually.

That flow has changed recently. @ClaudeDevs is Anthropic's official account. The account announced the new feature.

It's about the Claude Code desktop app. There's a new checkbox. It's called auto-continue.

One Checkbox, and You're Done

Using it is simple. Turn on the checkbox in the desktop app.

Even if you hit the limit, your request stays in place. The moment the limit resets, it continues the work automatically. You don't need to log back in.

You Watched the Reset Time Like Setting an Alarm

Until now, users had to keep track of the reset time themselves. They set alarms or logged in periodically.

When the limit hit in the middle of a long task, the flow broke. You had to figure out the context again and pick it up. This update removes that repetitive work.

For Long Jobs, Turn It On and Walk Away

It's useful for long tasks like extended refactoring or large-scale code generation. You can switch on the checkbox and step away. When the limit lifts, it carries on by itself.

That said, you should always review what it produces after continuing automatically. Checking comes first, not leaving it unattended.

Terminal CLI Support Still Needs Confirmation

Japanese tech outlets also reported this feature. They are PC Watch and forest.watch. The wording matched the original.

However, Anthropic's official documentation is different. There is no separate entry in the changelog or support docs. Whether it applies to the Claude Code CLI has also not been confirmed.

A related request has been filed as a GitHub issue. Whether it will be officially adopted is still not officially confirmable. If you use the desktop app, start by checking where the checkbox is now.

Frequently asked questions

What is the key point of “One Checkbox, and You're Done”?
Using it is simple. Turn on the checkbox in the desktop app. Even if you hit the limit, your request stays in place. The moment the limit resets, it continues the work automatically . You don't need to log back in.
What is the key point of “You Watched the Reset Time Like Setting an Alarm”?
Until now, users had to keep track of the reset time themselves. They set alarms or logged in periodically. When the limit hit in the middle of a long task, the flow broke. You had to figure out the context again and pick it up. This update removes that repetitive work.
What is the key point of “For Long Jobs, Turn It On and Walk Away”?
It's useful for long tasks like extended refactoring or large-scale code generation. You can switch on the checkbox and step away. When the limit lifts, it carries on by itself. That said, you should always review what it produces after continuing automatically. Checking comes first, not leaving it unattended .

Related articles