SilverleafSolutions

Codex usage limits are back to 5x/20x vs Plus. Here are some tips to stretch your quota: -Always use plan mode before allowing Codex to do work -Disable Codex memories in Personalization settings -Turn off noisy build warnings -Keep AGENTS.md small -Disable unused MCPs -Use smallest model/low reasoning for routine edits

@thsottiaux I'm a hoarder by nature. No different with resets. They make me plan to do a bunch of work in a short period of time to make the most use of a reset. I really don't want to waste it!

Now that you can bank usage resets in Codex. Are you a hoarder or do you use them without breaking a sweat? How do you think about them?

@thsottiaux I've had a number of problems with the app (UAC prompt spam, can't open Chrome, etc.). Would be awesome to have a repair process for the Windows app.

What should we improve in the Codex app. What's not delightful?

@housecor Anyone who has worked with AI agents knows they need constant checks. They make hidden assumptions, don't follow existing standards, or just plain forget requirements constantly. Not ready for autonomy yet.

The big divide in software development in 2026: Do you do human code reviews? If you do, then "loops", “away from keyboard”, and "multiple concurrent agents" are mostly useless. Why? Because these practices just generate a backlog of code to review faster than humans can review it. Human reviews are the bottleneck, not code generation.

@gdb I can do 100x more engineering than I could do before. More far-reaching tasks. Architecture that would have taken me hours or days, done in minutes. Tests I never would have written. As a creator/problem-solver first, coder second, I love this new world!

software engineering is so different now. hard to remember what it was like even 6 months ago.

An undisciplined software engineer using AI can be like watering a garden with a fire hose. Software engineering is a craft, a skill. Learn about the craft, so you can better wield the superpower of AI!

@MatthewBerman Yeah. They've got time. GPT 5.5 works so well, I'm not too bummed. But the part of me that loves cool, new tech is bummed though. 😁

so there's no incentive for OpenAI to release their next model right now, right?

Happy Father's Day to me! 🤤🥵 @PCBeefJerky's jerky is top notch and these spicy flavors are 🔥

@thsottiaux Why the decision to not allow editing of any text files from Codex itself?

Hi, I'm Tibo and I just discovered Codex. AMA

A point of advice, if I may be so bold: rest. I know: it's tempting to keep AI agents running at all times. I am often the blocker. But this leads to constant distraction. Not good. Be present with your family and friends. AI agents can wait!

@thsottiaux One thing no one can deny: you make your customers feel informed and excited about coding! I love the way you interact so openly on X, and the resets were already fun, but now just absolutely awesome that I can reset when I need to do more work. Great job!

Heard your (amusing) feedback that it was at times annoying to receive a reset of your Codex usage without warning. Next time we press the button you will get to choose when it actually applies. Happy codexing.

@tonysimons_ @thsottiaux Same. It feels like quota burn is around 30% more than even last week.

@thsottiaux Feels like we saw a massive reduction in usage limits today too, if we’re being honest.

Can confirm we saw a strong spike in growth of token consumption for Codex over last 48 hours. Unusual when we don't launch something.

@caps_raunak AI is a dream for creative people who don't want to troubleshoot for 3 days straight on a stupid problem. 😆

Bro, I'm so fucking tired of this AI shit. You pour 8–10 years into mastering a craft late nights, brutal feedback, real scars from real failures. Then some model trained on millions of stolen data (including yours) spits out better work in 4 seconds flat. And the replies? "This is amazing Future is here" We're not progressing. We're enthusiastically building the machine that makes our entire skillset worthless and calling it innovation.

@robertcourson This competition is great though. It will push both companies to do their best. Or at least try.

at this point I don't want GPT 5.6 - it needs to be GPT 6. No way Anthropic has completely blown past them like this.. 3 models in 2 months and Fable is not even their best model. Feels like anthropic ruined OpenAI's whole model roadmap/release plan.

@TimJayas Settle down there, Tim. 😁 x.com/BHolmesDev/sta…

Codex is straight up COOKED this week 💀 Even GPT-5.6 can't compete with Claude Mythos

@thsottiaux Quota limits are too low for the normal folk to do this yet.

Anyone writing nested loops yet?

@steipete I agree... if one has essentially unlimited quota, but this simply isn't realistic for most folks. I wonder how long you could have Codex loop on a goal on the $100 or $200/mo plans before you run out of quota...

Here’s your monthly reminder that you shouldn’t be prompting coding agents anymore. You should be designing loops that prompt your agents.

@gdb Since the $100/mo plan also allows ChatGPT Pro usage, I run Pro chats in the ChatGPT app, so it doesn't use my Codex quota. I would love to use Codex for everything, but low quota means I simply can't.

Whenever I don’t use codex for a task, I ask myself why and usually realize that there’s some missing context, I needed to write a skill, or I just didn’t think to use it. Rarely is it because the task is outside of the capabilities of the model. Overhang right now feels large.

If I see Codex use the term "smoke test" one more time.....

I didn't get into software to code. I got into software because I love building and creating. Code is a means to an end for me. If I don't have to code and get to see finished products quickly, huge win! This is the most fun I've ever had!

how much of your code is actually written by agents now?

Can't get your Azure Function App to build via the normal GitHub action due to "Error: Repository access blocked"? I ran into this tonight. Use this yml template to build it via zip instead. Use AI to customize it for your software. ``` name: Build and deploy Azure Function App via zip on: workflow_dispatch: env: AZURE_FUNCTIONAPP_NAME: 'FUNCTION_APP_NAME_HERE' AZURE_FUNCTIONAPP_PACKAGE_PATH: 'PACKAGE_PATH_HERE' AZURE_FUNCTIONAPP_RESOURCE_GROUP: 'RESOURCE_GROUP_NAME_HERE' DOTNET_VERSION: '10.0.x' jobs: build-and-deploy: runs-on: windows-latest permissions: id-token: write contents: read steps: - name: Checkout GitHub Action uses: actions/checkout@v4 with: submodules: true token: ${{ secrets.GH_PAT }} - name: Setup DotNet ${{ env.DOTNET_VERSION }} Environment uses: actions/setup-dotnet@v4 with: dotnet-version: ${{ env.DOTNET_VERSION }} - name: Publish Function App shell: pwsh run: | dotnet publish "./${{ env.AZURE_FUNCTIONAPP_PACKAGE_PATH }}" --configuration Release --output "./publish" - name: Package Function App shell: pwsh run: | Compress-Archive -Path "./publish/*" -DestinationPath "./functionapp.zip" -Force - name: Login to Azure uses: azure/login@v3 with: client-id: ${{ secrets.AZUREAPPSERVICE_CLIENTID }} tenant-id: ${{ secrets.AZUREAPPSERVICE_TENANTID }} subscription-id: ${{ secrets.AZUREAPPSERVICE_SUBSCRIPTIONID }} - name: Deploy Function App Zip shell: pwsh run: | az functionapp deployment source config-zip ` --resource-group "${{ env.AZURE_FUNCTIONAPP_RESOURCE_GROUP }}" ` --name "${{ env.AZURE_FUNCTIONAPP_NAME }}" ` --src "./functionapp.zip ```