You know that feeling when you send a prompt and immediately realize you forgot something? Or Claude goes off in a direction you didn’t want, and now you’re three messages deep trying to course correct?
Yeah. Turns out there’s a better way.
Just hit Escape twice
Double-tap Escape in Claude Code and you can jump back in your conversation history, edit your previous prompt, and try again. No new message. No “actually, what I meant was…” No watching your context window fill up with corrections.
You just… fix it and go.
Bonus: single-tap Escape interrupts Claude mid-thought while preserving your context. Handy when you see it heading the wrong direction and want to course correct before it finishes.
I’ve been using Claude Code for months and somehow missed both of these. Found them buried in Anthropic’s best practices guide1 under “course correct early and often.”
Why it matters
Every message you send adds to the context. When you’re deep in a coding session, that context is precious. Sending correction after correction is like having a conversation where you keep interrupting yourself.
Double-tap Escape lets you actually edit the thought instead of piling on addendums.
It’s a tiny thing. But tiny things compound.
Anyway, that’s it. Go try it.
Links:
- Claude Code’s hidden conversation history (and how to actually use it) - Another quick tip for navigating Claude Code
- My updated Claude Code workflow - The full breakdown of how I’m using Claude Code