I noticed when chatting with the new agent with Claude 3.7 that it keeps adding and editing code for queries where I just ask it questions or ideas.
Anyone else notice this?
I even tried adding a rule to tell it to always first plan but it doesn’t follow it.
3 Likes
What always works for me is putting at the end: DO NOT CODE ANYTHING, JUST GIVE INFO. It has never ignored this.
right but i dont want to have to do it in every session 
Yes, I think its a problem with Sonnet 3.7.
I will disable it for a while.
Todays cursor is nuts.
This is interesting, I was literally about to post here or on X that my last 20 responses from Cursor Claude 3.7 all started with verbatim “You’re absolutely right, …”
1 Like
Tinker with the cursor rules (the general ones that are applied to all chats) to instruct the model not to code anything when you say [whatever keywords you want] or something in this spirit.
1 Like
After disabling 3.7 things are MUCH better.
Hmm yeah I think 3.7 needs to get some guidance. By default it seems to try too hard to please you by fixing all sort of things and running the dev server and whatever. It’s a bit wild.
I used 3.7 on the first day it was available before the rate limiting mess and found it doing what’s everyone is complaining about in this thread. It would go off the rails, ignore some commands, take initiative that doesn’t need to be taken, etc. Yesterday, I reverted back to Sonnet 3.5 and smoothly coded through massive parts of my project I’ve been working on. I think, as of now, until kinks are worked out and usage drops a bit with Claude 3.7, I prefer Claude 3.5. I’m curious if the lack of usage of Clause 3.5 all of a sudden is improving the model, lol.
use notepad “DO NOT CODE ANYTHING, JUST GIVE INFO OR UNTIL I TELL YOU TO CODE”
I noticed that you can change between ‘Modes’ mid-chat. So you can ask some question in ‘Ask’ mode and then switch back to Agent. This way forces a text-only response. It doesn’t access tools though so it won’t retrieve additional context (or so it seems). Maybe Edit mode is the answer here.
Either way, I’ve been finding 3.7 a bit harder to use so far. I might go back to 3.5 as well. I haven’t tinkered with my rules just yet though.
Yes, i have a shortcode for it. - aoni - analysis only, no implementation, which I have added to the rules.
So, aoni at end of messages. 
Ive had similar issues. 3.7 can end up going in circles more then 3.5
I have reverted to 3.5 for now.