[BUG] Missing visual feedback for thinking trigger words in Claude Code 2.0

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

In Claude Code 1.x, when typing thinking trigger words (think, think hard, think harder, ultrathink), the trigger word would be visually highlighted/colored to provide feedback that the thinking mode was being activated. This visual cue helped users confirm they had typed the trigger word correctly and that Claude Code recognized it.

As of Claude Code 2.0.2, this visual feedback has been removed for think, think hard, and think harder. Only ultrathink still shows the colored highlighting.

What Should Happen?

All thinking trigger words (think, think hard, think harder, ultrathink) should be visually highlighted or colored when typed, providing immediate feedback that:

• The trigger word was typed correctly
• - Claude Code recognized the thinking mode trigger
• - - The appropriate thinking level will be applied

This was the behavior in Claude Code 1.x and is still present for ultrathink.

Error Messages/Logs

N/A - This is a visual feedback issue with no error messages. The functionality still works, but the visual highlighting is missing.

Steps to Reproduce

1. Open Claude Code 2.0.2
2. 2. Type a prompt containing think (e.g., "think about how to optimize this function")
3. - Observe: The word "think" appears as plain text with no visual highlighting
4. 3. Type a prompt containing think hard (e.g., "think hard about the best architecture")
5. - Observe: The phrase "think hard" appears as plain text with no visual highlighting
6. 4. Type a prompt containing think harder (e.g., "think harder about edge cases")
7. - Observe: The phrase "think harder" appears as plain text with no visual highlighting
8. 5. Type a prompt containing ultrathink (e.g., "ultrathink about this complex problem")
9. - Observe: The word "ultrathink" IS highlighted/colored (this works correctly)

Expected: All four trigger words should show visual highlighting
Actual: Only ultrathink shows highlighting; think, think hard, and think harder do not

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

1.x (Claude Code 1.x)

Claude Code Version

2.0.2

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Warp

Additional Information

Impact

Severity: Low-Medium
User Experience: This reduces discoverability of the thinking triggers feature and removes helpful visual feedback that aided correct usage.

UX Context

The visual feedback served an important purpose:

1. Immediate confirmation - Users could see their trigger was recognized without waiting for the response
2. 2. Typo detection - If the word didn't highlight, users knew they had a typo (e.g., "thnik" instead of "think")
3. 3. Consistency - All thinking levels had the same visual treatment, making the feature discoverable and predictable

The fact that ultrathink still has this coloring suggests this may have been an unintentional regression rather than a deliberate design decision.

Suggested Solution

Restore the colored highlighting for all thinking trigger words (think, think hard, think harder) to match the current behavior of ultrathink.

Related Issues

• #7668 - Configuration and Documentation for Thinking Mode (discusses thinking trigger keywords but not the visual feedback)