Huge Thanks to the Cursor Team – You Nailed It!

Feedback
1

I just wanted to take a moment to thank the Cursor team — especially the incredible devs and pricing folks — for this latest update. You absolutely nailed it.

I’ve been using Cursor for around six months now. I started back on the free tier, eventually subscribed, and I always ran into the 500 agent request limit within just two weeks. It was frustrating at times — especially when doing freelance work and needing to move fast.

I’ll admit: at one point I even considered alternative methods or plataforms (which I really don’t recommend anyone do, ever) just to get things done and try to recover the money I had invested.

Here in Brazil, and in many other countries where people don’t earn in USD, the pricing can feel scary — our minimum wage is about $200 USD, so paying $20 a month for a tool isn’t like the price of a coffee, it’s closer to the price of a coffee machine. But even so, I always saw real value in Cursor. It made me more productive, more focused, and more capable.

So when I saw the update — that the Pro plan now works with just rate limits, not hard usage caps — I literally felt nothing but gratitude.

This is the kind of change that makes a real difference. You listened to the community and delivered something powerful, fair, and respectful of your users.

So again: thank you, from the bottom of my heart, to the Cursor dev team and pricing team. You’ve built something truly special — and now you’ve made it even better.

thats it, thx <3

2 Likes
2

Hi @davedev and welcome to Cursor Forum :slight_smile:

Thats a great first post and thank you for the positive feedback. I will pass the thanks certainly directly to the Cursor Team as well. :heart_with_ribbon:

Could you share a bit what you are doing with Cursor, what models you use and how it works for you?

Many in the forum are interested to see what others do and what works

3

Hi T1000! Thanks for the warm welcome :slight_smile:

I’ve been using Cursor mostly for refining business logic through structured prompt engineering. I use cursorrules a lot to maintain a consistent format and work through complex reasoning.

For code generation, I usually prefer Claude — especially since Sonnet 3.5 (non-thinking). The experience has been great. Claude 4, in particular, is just perfect. Simple as that.

I still use GPT models like o1 or o3 whenever I need deep structural thinking or to organize ideas. They’re super helpful for defining business rules and architecture before diving into code. Once that’s clear, Claude can handle most of the implementation — even repetitive stuff you’d otherwise have to write manually.

Cursor’s @search is also a game changer when I need to avoid digging through documentation manually.

At the end of a session, I always ask Claude for a checklist of what we’ve done. That gives me a nice test plan or validation checklist to go through step by step.

If you’re starting with cursor, here are my recommendations:
Always enable terminal MySQL\YourDB read-only mode to block dangerous commands and set (by cursorrules and block commands directly)

NO_TINKER (using php for exeample) to prevent accidental DB modifications.

For cleaner code, enforce type over interface and enable ESLINT autocompliance.

Optimize workflows by requiring frontend hooks separation and unified API service calls (like using api.ts instead of direct axios).

Finally, assume a Linux environment (or use wsl, but commom just use linux, its just better at all, it will allow you to develop more commands and install dependencies directly on cursor, thats just amazing! Windows Powershell has limits on that) and require migrations via terminal only. Start with 3-5 core rules and expand as you spot workflow patterns!

Hope this helps someone!

1 Like
4

Greetings to Brazil :slight_smile:

2 Likes
5

wow thats a lot of good advice you shared. Thank you :slight_smile: really appreciate it.

My experience is similar in that Claude Sonnet (3.5 and now 4) can do so much of the implementation very well. I guess thats one reason why its very popular.

How did you enable terminal MySQL read-only mode? By allowing selects but not inserts/deletes or prefixing conection with read only mode on?

Im sure this will be very helpful to many!

6

I’m not currently working on that project, but you can really do that on Auto-Run Rules:

{B03F2A21-9450-4881-9E4B-93FB87374CDC}
{B03F2A21-9450-4881-9E4B-93FB87374CDC}650×606 15.2 KB

And to confirm that, you can add on the top lines of cursorrules/userrules

******** WHEN ACESSING IN TERMINAL = ITS FORBIDDEN USING MYSQL CONNECTIONS TO “DROP” “ALTER” “UPDATE” “SET”!!! - THE ONLY FAIR USE FOR MYSQL CONNECTIONS IN TERMINAL ITS TO READ-ONLY, THAT MEANS SELECTS, JOINS AND ETC THAT IS INSIDE THE CONTEXT OF READING! OBVIUSLY IN MIGRATIONS OR CONSTRUCTING THE APPLICATION, WE WILL HAVE FUNCTIONS LIKE DELETE AND UPDATE, BUT ITS CODE! NEVER IN TERMINAL!*********

Another good thing here: if you are working in a multi-space project (frontend+backend, for exeample) always use something like this on userrules:

FRONTEND = ~/projetos/php/aistenlab/yourfolderfrontend/

BACKEND = ~/projetos/php/aistenlab/yourfolderbackend/

why = oftenly you will refer to just (do that in frontend) this can help ai to find directories quickly

this will REALLY save your life haha

It worked really amazing when:

  • Made a Feature
  • Test directly via curl in the terminal
  • verify in the databse with terminal commands if the data is now updated!
  • Repeat

this flow is very productive!

1 Like
7

Nice :slight_smile: I’m usually allowing Tinker but give Agent only access to a separate development environment and it may call only commends with --env=development (or testing) that has separate DB for Agent only.

1 Like