populate shell history?

[benwhalley]

This is really cool, but I can't use the up arrow to re-run a command. Is it possible to populate the terminal history to enable this, or just enter the command to be used to the terminal without running it so that actual execution happens normally rather than inside wz?

[searchableguy]

That's a good suggestion. I will need to check how that would work with child_process.

I was already thinking of feeding LLM the terminal history to chain commands via natural language better.

[searchableguy]

@benwhalley So this is sort of possible but complex. The issue is adding to the user history will differ based on the shell user is using. I can add shell specific code to add history but I think we can only cover the major ones such as powershell, zsh, bash, etc.

[austinweisgrau]

Agree that this would be helpful! I'm using bash.

Find myself wanting to run the same command several times and the "up arrow, enter" muscle memory keeps getting me as I resend the same slow gpt network call.