Feedback - Edit Box Functionality in Claude Cade and Commands Access
I’d like to raise a feedback around the core edit box experience in Claude Code. While it serves its base purpose well, it's currently limiting for deeper or more complex workflows. Below are some key issues and suggestions that would make a big difference in daily usability:
---
1. Limited Text Editing Capabilities
- No Mouse Cursor Placement: If you’ve written a long prompt or pasted in extensive content, there's no way to jump the cursor to a specific point using the mouse. You’re restricted to using arrow keys to manually navigate through large blocks of text to edit it.
- No Text Selection for Deletion: You can't highlight and delete a specific chunk of text. Standard editing tasks like cutting or trimming sections, or moving text around become unnecessarily cumbersome.
These feel like very basic expectations from any modern editor interface and would massively improve prompt iteration.
---
2. Slash/Question Mark Command Invocation Limitations
- No Inline Slash Triggering: If you've already written a multi-line prompt and then realize you want to invoke a command (e.g.
/user:ask,/project:context_prime, etc.), you’re out of luck. The slash menu only works if the/is the first character in the box. - Command Flexibility Needed Mid-Prompt: Sometimes we build a detailed prompt, forget to invoke a command at the beginning, and later realize it should’ve started with a slash. At that point, even if we navigate back to the beginning and insert a
/, the system no longer accepts it as a valid command invocation. It only works if/is the very first character typed in a fresh prompt, which is painful.
---
3. Undo and Redo
Be great quality of life update to get Undo and Redo functionality in the core edit text box
---
Final Note: Elevating the Core Editor Experience
This edit field is the central interface for nearly every interaction with Claude Code—your primary canvas for prompt construction, your command line for creative logic, the nerve center of the tool, and the gateway through which all task instructions flow. And yet, in its current form, it feels neglected—lacking the basic affordances we’d expect from any serious development environment. Investing in what might be considered core interaction infrastructure—things like selection, cursor precision, slash invocation flexibility, and undo/redo—would go a long way. These aren’t wishlist extras; they’re foundational to making Claude Code feel robust, comfortable, and ready for real work.
This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗