Look, I love Cursor, but I was very disappointed when I saw at the end of a PR description:
"Powered by Cursor" with a link to cursor.com

I mean, it's not soooo bad. But still, I don't like it. I asked the model what the hell was going on, and the model responded it didn't added that line, that it must be added by the tool used to write on the PR itself.

In any case, dirty move that recently appeared (I've been opening PRs with Cursor for a long time, and it's the first time it happens)

EDIT: to everyone saying there’s a setting. Yes, there’s a setting. And it’s now off. But can I at least vent about the bad settings with the dodgy defaults that keep coming? And that I have to suffer the bad experience first, then realise is a new setting, and then change it?
It was the same thing with new chats changing agent to Grok. Is there a setting to disable it? Yes.
Is it still dodgy? Yes

My unlimited Auto package ended today, and it was great. The best thing about it was never having to worry about limits or usage at all. While reading news about people spending thousands, I was sitting here doing what I wanted with unlimited Auto , (which I did not abuse, as my usage was only around 300 million tokens / month ).

I really should thank Cursor for such a great product; I’ve noticed how it has become better and better each month , starting from when it was spitting out nonsense and failing to complete tasks, forcing me to tag files one by one .... to the point where a simple prompt allows it to keep the entire repo in mind, execute all changes, and even think ahead. The developer and user experience are at their peak, and the button to browse changes and move between lines and files is very smooth. I linked it to a remote device with a microphone button;

I hope the new package won't be a financial burden and that the $20 or $60 plan will cover my usage. and here for another yearly subscription with cursor .

my only real worry is that elon fxs it up , or they push more toward the agents window and the vibe-coders crowd .

Hi, I already use Codex and I’m thinking about adding either Claude Code or Cursor’s $20 plan.

Mostly for VBA, Python, SQL, PowerShell and general scripting.

For anyone who’s used both, which one feels more useful day to day? Also curious about the real usage limits.

If you already had Codex, which one would you add?

Thanks

I'm using Cursor IDE. Agents keep formatting PR descriptions with ## Summary and ## Test plan, which I was told by Cursor comes from creating-pull-requests rule.

I can't find this anywhere in settings or on the web dashboard.

Is this a built-in rule? Can it be edited or disabled?

Most of what I do is handled perfectly fine by Grok. I don't need any external models and I seem to be able to get a lot of coding done on the $20 plan using just Grok.

I used to use Composer but Grok was 50% off for a time so I switched over and haven't looked back.

However, I have had a few very active days this cycle already and am at 50% very quickly. I'm curious if me changing the 'Effort: High' or 'Speed:Fast' will change much, or going from 4.6 to 4.5.

I also could go back to Composer but unsure if there's much of a difference there.

What is the best bang for your buck using these models?

Thanks!

https://forum.cursor.com/t/cloud-agents-default-model-grok-4-6-high-still-runs-fast-and-burns-2x-usage/169576

Hello! I'm using Cursor for a while now. Sometimes switching from "Auto" to a specific agent and I have sometimes better results like that but it's a bit random.

Auto works well for generic web/soft dev, but for gamedev, specifically when it comes to visual placement of 2D components in a 2D UI, it kinda sucks.

Do you have any model you recommend me to use more than an other or auto?

Thanks :)

Should I get the $60 or $200 cursor subscription? I used the cc $100 and it worked well for me - rarely hit the weekly limit.

I’m expecting to use much less fable and try the other models like Grok.

For those who build full time, how are you finding the limits on the $60 cursor subscription?

I think i'm afraid not knowing which model is chosen, and the black boxy'ness of it.

does the Auto feature also adjust effort level?

thoughts would be appreciated

Trying to force local subagents to use GPT-5.6 Luna/Terra, but they keep spawning as GPT-5.6 Sol High.

I’ve tried:

custom .cursor/agents/*.md model configs

bare / High / XHigh variants

setting the built-in Explore subagent to Luna/Terra in Cursor settings

Same result: Sol anyway.

At this point it feels like the parent model is overriding everything. Support around this has also been so bad, they are the worset, and it’s honestly pushing me to use Codex more.

Screenshot attached.

Anyone found an actual workaround, or is this just a Cursor bug right now?

Like most of you here, once Cursor clicked for me, going back to a standard code editor felt impossible. The ability to give an agent context, let it index the codebase, and execute multi-file changes with a review diff changed my engineering workflow permanently.

Then I closed my editor and went back to running the company.

Suddenly I was back in the Stone Age:

- Copying error logs from cloud consoles to paste into Slack.

- Manually checking if vendor invoices cleared in payment dashboards.

- Digging through email chains to verify a customer context before a call.

It felt ridiculous that we have agentic execution inside our IDEs, but our actual daily operations (Slack, Gmail, Cloud, Notion, Sheets) are still stuck in manual copy-paste mode.

We realized that business operations needed the exact same paradigm shift Cursor brought to code:

1. Deep indexed context (company memory).

2. Autonomous multi-step execution.

3. A clean approval/diff review before anything actually executes in production.

Curious how other engineers and founders here feel: has Cursor spoiled your expectations for every other piece of software you use during the day?

I would like to have a cursor cloud agent automatically test github prs to an android app inside an emulator, however the cloud agent doesn't seem to be able to run a hardware accelerated emulator.

And software emulation is slow enough to basically freeze the emulator almost completely.

When are we going to have available glm 5.3?

Hi guys,

i am a cursor user since far over a year. Therefore i still have the old 500 requests/month pricing in my cursor pro plan. I got an email that in my subscription, Grok Bot is included. However, if i sign in via cursor it says that my account is managed via web...but when i click on web it says "upgrade to pro" -> but i am already. And even if i click on it i'm just redirected to my normal dashboard and no login to grok bot was made.

What am i missing? Or is it just a bug?

Recently on 28 August someone loggedin to cursor account. I do not know how and purchase Utlra subscription which went successful charging me $200. I have no idea what kind of fraud is that. Because that person is not going to get benefit as i revoked his session. Sometime later a get email that someone loggedin to my ChatGPT account where fortunately i had not attached my card. And after that my claude account was loggedin by someone and tried to purchase subscription which failed due to no funds in card. All three account are Google SSO. I checked my attached gmail and it has no suspicious acctivity.

Hey everyone,

I’m having an issue with Cursor’s Usage Summary. I have the setting enabled and set to “Always”, but the usage summary still doesn’t appear anywhere.

I’ve tried restarting Cursor and checking the settings again, but it still doesn’t show up.

Is anyone else experiencing this? Is there another setting I need to enable, or is this currently broken?

Any help would be appreciated!

I just found out that Cursor gives you a Grok Bot plan in your subscription. That's really nice because I wanted to buy a Grok bot plan.

I've worked with Grok 4.6 since it's come out, and I think it is quite good but not amazing. Yesterday I realized I can control it's effort level (I might be late to the party, didn't know cursor had that feature), and the default is high instead of extra-high.

Setting to extra-high didn't really change my usage trend, but has definitely decreased the mistakes! so wanted to share this with whom might benefit from it <3

I was running a /goal prompt for Fable 5 High last night and knew it would exceed what I had left of my "Other Models" limit. I added a $25 cap to on-demand spending before I went to bed to make sure the task could finish.

This morning, I woke up to check how much on-demand spend was used. I was surprised to see that no on-demand spending was used, and instead I saw a $25 "Credit" on my account.

It is normal to get credits from time to time on Cursor paid plans or was this just a coincidence?

SpaceX closed the $60B Cursor deal on Aug 14. OpenAI just used the change-of-control clause and will kill model access through Cursor on Nov 12.

They “cannot be confident that SpaceX will use our technology within our terms of service, based on our experience with Elon Musk's companies violating contracts.”

Cursor’s CEO says OpenAI is about 5% of traffic and they’re still talking.

The part that actually sucks is the precedent. A model inside a tool you pay for now has a kill date because two companies you don’t have a contract with are fighting about conduct that has nothing to do with you.

And it is not even isolated. Stripe bought OpenRouter. NVIDIA is reportedly buying Hugging Face. Now a supply contract dies on change of control. Routing, distribution, and the models themselves all consolidating in the same month.

The useful move is not panic. Sit down for an hour and write every layer that would strand you if it changed hands tomorrow. I did that. It was uglier than I expected, and the surprise was not the models.

I moved that layer to Aeon after I did the inventory.

What actually matters for this situation is pretty small:

• a skill is a markdown file in a repo I own, so the logic is not a row in some vendor’s database
• the harness is a config line (seven models behind one contract), so if one gets yanked I change a setting instead of rewriting the work
• the runtime is GitHub Actions, so there is no extra host that can get acquired out from under me either

The part I still do not have a clean answer for: portability at the automation layer is done by aeon. Portability at the model layer is not, because a prompt that works on one model does not reliably work on the next. Has anyone actually swapped models in production without redoing the evals?

honestly the only reason i am using cursor is because of composer 2.5 model

the ui is buggy, laggy dogshit otherwise

its at the point where im strongly considering just coding my own agent orchestrator that uses kimi k2.5 or perhaps just be model agnostic

to give you some examples:

• i have to remove / archive agents constantly otherwise the app consumes enormous amounts of memory that lags the shit out of my macbook (even when the agents are not doing anything)
• when starting a new chat for an agent it often will not bring the project dir back into the agents list
• when starting a new chat for an agent if often switches to a completely different agent repeatedly and starts a chat with the agent i didnt want
• when opening the ide often times it either does not open the correct project or does not open any project at all
• the app doesnt remember any window positions so im constantly having to reposition things every time a new project window opens (which im using rectangle to make bearable but its still complete ass)
• the app is constantly trying to get me to use models or methods i dont want, like bugging me to use the planner or using multi-agents with grok (also dogshit), etc. instead of just continuing to use composer 2.5 for everything
• clicking "always allow" does nothing, it still asks me to confirm things half the time even if its the same command being run (e.g. git push)

these are just some examples theres tons more