feat: let user tweak architect prompt before passing to editor-model #3090
+55
−24
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
closes issue 3085 : #3085 (comment)
This is just a demonstration for those who want to test it out... despite there being room for improvement, I use this happily in my local setup I thought I'd share it in case anyone wants to use it.
There is something I'm not 100% sure of yet: the intended design is that the resulting context doesn't contain any 'old content' from the architecture model's first try. The context should only contain the manually edited architectural prompt to then hand over to the editor model. In theory this is what my PR does but I'm not familiar enough with Aider to test it.
Another thing I did here is that I have cleaned up how the kwargs dictionary was being initiated, and deleted an unused 'validate' function.
I should also note that I haven't ran pytest for this change. Finally, I haven't given the option for users to configure which editor they want to use, so this is also a to-do for this feature, however they can over-ride this by setting EDITOR in their ~/.bashrc or ~/.zshrc file.
Any comments are appreciated!