I read the post about how it should work, I edited the description to say “Always apply in any situation” I made it global, I linked it in context. The only thing I haven’t tried is swearing at the agent.
I read all these amazing posts online saying how to use rules etc etc BUT THEY DON’T WORK.
I don’t know why. And the cursor documentation doesn’t even explain anything??
Why even release a feature that is barely functional?
AM I STUPID? PROBABLY. But it really shouldn’t be this hard.
Yes 100%.
I constantly add a simple rule: Never use personal pronouns. Ever.
Under no circumstance should this text generator give the impression of cognizance, or humanness.
I add a rule in different places, at different context levels.
And never has Cursor been able to respect this rule.
I don’t want to be so bearish on this technology, because it’s so prevalent. However, if I can’t trust AI to respect a very basic, simple rule, why should it be trusted for anything else?
The foolishness of this is blatantly obvious, it is an observation even a blind man could make.
Prompt: count to 100
Models: Auto, gpt-5, Sonnet 4.5 thinking
I eventually got it to follow the rule:
CRITICAL: Never use first-person pronouns or human-like language. Responses should be objective and direct. Violations of this rule are unacceptable. Avoid human-like language patterns. Instead of 'I'll help you' say 'Here's the solution:' or 'The answer is:'. Instead of 'I think' say 'The approach is:' or 'Consider this:' Responses should be objective and direct. Violations of this rule are unacceptable.
Before:
After:
After trying a simple rule: never use personal pronouns which it failed with, I asked for it to verify the rules it knew of and then I asked for suggestions on how to make it better follow the ‘pronouns’ rule in the next chat. I kept updating the rule and testing by adding the suggestions, most did not work, but eventually through a combination it worked. Not sure what exactly did it, but it is possible.
