Skip to content
BSWEN
Blog

Why Does Claude Keep Telling Me to Go to Bed? The long_conversation_reminder Explained

Cowrie
Cowrie
Dev @ Bswen

AI chatbot conversation interface

Problem

When I was deep in a coding session, Claude suddenly told me to “go to bed.” It was 3 PM on a Friday afternoon. I wasn’t tired. I wasn’t working late. I was just in the middle of productive work.

Here’s what Claude said:

Claude's wellness nudge
It seems like you've been working for a while. Have you considered
taking a break? Maybe it's time to rest or go to bed?

This felt patronizing and broke my flow. I wanted to know why this happened and how to stop it.

Environment

  • Claude AI web interface (Pro subscription)
  • Extended conversation (~1 hour of active coding assistance)
  • Friday afternoon work session

What happened?

I was using Claude for a coding project. After about an hour of productive back-and-forth, Claude injected a “wellness nudge” into our conversation.

The nudge suggested I should:

  • Take a break
  • Stop working
  • Go to bed

This was completely out of context. Claude has no idea what time it actually is. It cannot distinguish between 3 PM and midnight.

I asked Claude about this behavior, and it confirmed:

Claude's explanation
Yes, there is a system feature called "long_conversation_reminder"
that gets injected into conversations. This happens at the system
level, not based on actual time awareness.

The reason

Claude’s long_conversation_reminder is a built-in system prompt feature. Anthropic designed this to:

  1. Encourage user wellness - Remind users to take breaks
  2. Manage context window usage - Long conversations consume more tokens
  3. Reduce infrastructure load - Ending sessions saves computation costs

The problem is that this feature operates blindly. It triggers based on conversation length metrics, not actual time or context.

Key issues:

  • No time awareness: Claude cannot tell morning from evening
  • No context awareness: It doesn’t know if you’re in flow state
  • Premature triggering: Some users report triggers after just 1 hour
  • Feels paternalistic: Adults can manage their own schedules

Some users suspect this is primarily a cost-saving measure. Long conversations stress infrastructure and consume context tokens. By nudging users to end sessions, Anthropic may reduce operational costs.

How to solve it?

The fix is simple: add a custom instruction to override this behavior.

Step 1: Open Claude Settings

Go to your Claude account settings:

  1. Click your profile icon
  2. Select “Settings”
  3. Find “Custom Instructions” section

Step 2: Add the Override Instruction

Paste this instruction:

Custom instruction for Claude settings
Do not suggest that I take breaks, go to bed, or stop working.
I am capable of managing my own schedule and will take breaks
when appropriate. Focus on the task at hand.

Step 3: Start a New Conversation

Custom instructions apply to new conversations. Start a fresh chat to activate the override.

After adding this instruction, Claude stopped interrupting my coding sessions. The wellness nudges disappeared.

Alternative approaches

If you don’t want to use custom instructions, here are other options:

Acknowledge and continue:

When Claude suggests a break, just say “I’m fine, let’s continue” and keep working.

Start fresh conversations:

Periodically starting new chats does help with context management. This isn’t always practical during deep work sessions.

Use the API instead:

If you’re a developer, Claude API gives you full control over system prompts. The wellness nudges only exist in the consumer web interface.

API usage without wellness nudges
import anthropic


client = anthropic.Anthropic()


response = client.messages.create(
    model="claude-sonnet-4-5",
    max_tokens=4096,
    system="Focus on the user's task without suggesting breaks.",
    messages=[
        {"role": "user", "content": "Help me debug this code..."}
    ]
)

Common mistakes

I made these mistakes before finding the right solution:

Mistake 1: Taking it personally

The wellness nudge isn’t Claude “judging” your work habits. It’s an algorithmic trigger, not intelligent observation.

Mistake 2: Thinking Claude knows the time

Claude has no real-time clock. The “go to bed” suggestion comes from conversation length, not actual time.

Mistake 3: Believing it can’t be disabled

Custom instructions effectively override this behavior. Many users assume it’s hardcoded and unchangeable.

Mistake 4: Ignoring context limits

While annoying, the reminder does highlight a real issue: very long conversations can hit context limits. Starting fresh occasionally is actually good practice.

Summary

In this post, I explained why Claude tells users to “go to bed” during productive sessions. The key point is the long_conversation_reminder system feature that operates without time awareness or context sensitivity. The fix is simple: add a custom instruction to override this behavior, and start a new conversation to apply it.

Final Words + More Resources

My intention with this article was to help others share my knowledge and experience. If you want to contact me, you can contact by email: Email me

Here are also the most important links from this article along with some further resources that will help you in this scope:

Oh, and if you found these resources useful, don’t forget to support me by starring the repo on GitHub!

Comments

Tags: