Feature Request: Message Queue/Do Not Disturb Mode During Task Execution

Feature Request: Message Queue/Do Not Disturb Mode

Problem Statement

Currently, Claude Code processes incoming messages immediately, which can interrupt ongoing complex tasks. This interruption breaks the workflow and context, especially during multi-step operations like refactoring, debugging, or complex code generation.

Proposed Solution

Implement a message queuing system similar to Cursor IDE that:

1. Queue incoming messages when Claude Code is actively working on a task
2. Prevent interruptions during task execution
3. Process queued messages automatically after completing the current task
4. Provide visual feedback to users that their message is queued

User Experience Flow

Current behavior:

1. User sends message: "Refactor authentication module"
2. Claude Code starts working on refactoring
3. User sends another message: "Also update the documentation"
4. ❌ Claude Code interrupts current refactoring to process new message
5. Context is lost, work may be incomplete

Desired behavior:

1. User sends message: "Refactor authentication module"
2. Claude Code starts working on refactoring
3. User sends another message: "Also update the documentation"
4. ✅ Message is queued with visual indicator: "Message queued (1 pending)"
5. Claude Code completes refactoring without interruption
6. Claude Code automatically processes queued message: "Also update the documentation"

Configuration Options

Provide user control through settings:

{
  "messageQueue": {
    "enabled": true,
    "autoProcessQueue": true,
    "showQueueIndicator": true,
    "queueBehavior": "smart" // "always" | "smart" | "manual"
  }
}

Queue behavior modes:

• always: Always queue messages during task execution
• smart: Queue only during complex/multi-step tasks (AI determines)
• manual: User controls when to enable/disable queuing (e.g., /queue on)

Use Cases

1. Complex refactoring: User realizes additional requirements while Claude is refactoring
2. Multi-file changes: User wants to add more files to update without stopping current work
3. Iterative development: User has follow-up thoughts while Claude is implementing
4. Exploratory coding: User discovers related issues while Claude is working on the primary task

Benefits

• Improved workflow continuity: Tasks complete without interruption
• Better context preservation: Claude maintains focus on current task
• Enhanced user experience: Users can share thoughts without breaking flow
• Increased productivity: Reduces need to wait or carefully time messages

Reference Implementation

Cursor IDE implements this successfully by:

• Showing a queue indicator when messages are pending
• Completing current task before processing queued messages
• Maintaining context across queued message processing

Additional Considerations

• Cancel mechanism: Allow users to cancel queued messages if needed
• Priority queue: Option to mark urgent messages that should interrupt
• Queue limit: Prevent unbounded queue growth with configurable limits
• Queue persistence: Optionally persist queue across session restarts

Alternative Approaches

If full queuing is complex, consider:

1. Simple toggle: /pause and /resume commands to control message processing
2. Confirmation prompt: Ask user "Claude is working, queue this message? (y/n)"
3. Smart detection: Auto-detect when message is related and append to current task

Community Interest

This feature would benefit:

• Developers working on complex, multi-step tasks
• Teams using Claude Code for large refactoring projects
• Users who think iteratively and want to add context without disrupting work
• Anyone who values uninterrupted AI workflow execution

---

Environment:

• Claude Code: Latest version
• OS: Linux/macOS/Windows (affects all platforms)

Thank you for considering this feature request!