RooFix AI (@RooFixAI) header
RooFix AI (@RooFixAI) profile picture

This isn’t automation. It’s speed as a weapon for commercial roof repair sales. #RooFixAI

USA
Joined January 2021
Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)Photo by RooFix AI (@RooFixAI)
Called it. 😒
Get ready. The left is about to cause the #AIBubble to burst.
40
Get ready. The left is about to cause the #AIBubble to burst.
It’s a Brave New World
1
90
The shared usage notice landed late in Tibo’s 28-day series. Codex and Work already shared a pool; December expands it to all of ChatGPT. Many of the daily updates were resets rather than lasting capacity gains.
1
66
RooFix AI retweeted
ChatGPT web app now has a new hidden shared usage limit checkout disclaimer - "Starting in December, plan usage will have a shared limit across all of ChatGPT and Codex"
226
260
321
2,112
803,130
Even @claudeai is out here giving us BANKED resets that don’t affect our weekly reset schedule. Looking at you @thsottiaux 👀
63
RooFix AI retweeted
Replying to @thsottiaux
@thsottiaux I get what you're saying. However, it is clear that most everyone else is confused by your post. Here's what the official OpenAI docs say (correct me if I'm wrong): For US Pro users from OpenAI’s docs: ⚠️ Deeper/background work draws on a plan allowance for “deeper work,” with extended limits for the first month after launch. ✅ Conversations with your dot don’t count toward ChatGPT usage limits (and don’t draw on weekly usage). ⚠️ Tasks your dot starts or manages in Codex or ChatGPT Work still count toward those products’ normal usage limits. ✅ First dot is fully included in Pro at no extra cost and available 24/7. Official wording is consistent across the Introducing dots post, ChatGPT Learn docs, and Help Center. Tibo, you should really try to provide a bit more clarity and context when explaining things like this lol help.openai.com/en/articles/… openai.com/index/introducing… learn.chatgpt.com/docs/dots @OpenAI @OpenAIDevs @sama @MatthewBerman
2
2
145
This 👇 is a REAL Day 16/ update! Great job @OpenAI, @OpenAIDevs, and @thsottiaux 💪
Day 16/ You can now cache for 12h in the API and for logged in users across Codex & Work we keep cache forever and you don’t have any penalty for coming back to an old conversation or resuming after a few hours and are now guaranteed to hit the cache.
55
RooFix AI retweeted
just released the 2.0 version of the un-slop AI-looking apps skill a guy already told me he cancelled his $15/month mobbin sub after trying V2 i spent over a week rebuilding it based on your feedback new stuff: → native iPhone + Android designs → animated MP4 app demos → better UI copy, onboarding & paywalls → App Store + Google Play assets → much better design taste it also includes everything from V1 the original stays free, Pro is just €5 (still ridiculously underpriced) retweet + comment "V2" and i'll send you both free and paid version (follow so i can dm)
822
344
10
1,188
95,902
RooFix AI retweeted
Replying to @MatthewBerman
Sounds like Meta’s system prompt or system harness instructions leaked into the response. It’s essentially telling the model not to think about how it was developed. It reads like something you’d only tell the model if it tended to ask, think, or reason about its previous versions and had attempted to retrace its existence or evolution. Either way, I’m here for it lol 🤪🫣
1
1
2
280
RooFix AI retweeted
Replying to @MatthewBerman
@MatthewBerman - That’s a great idea, however, I’d take it one step further so you don’t have it looking at nonsense every day. 1. Create a ‘daily active ledger’ .md doc. 2. Then create a manually triggered skill `/ledger` that tells the agent to get the Codex session ‘ID’ and add it to the ledger. 3. Run ‘/ledger’ when you start or walk away from a session if you think it’s something worth monitoring. 4. Then add something like this to the beginning of your original prompt:👇 “Read the `daily-active-ledger.md`”
1
1
73
RooFix AI retweeted
Replying to @thsottiaux
Mobile create is distribution. Definitely not a meaningful CODEX improvement. Either way…**the control surface is still the gap** here. But if you actually wanted to improve dots: 1. Customizable dots-specific `AGENTS.md` and `SOUL.md`. 2. Inspectable, editable memories. Right now clearing a dot memory means deleting the dot. 3. Named sub-dots, not just Codex or Work tasks. A sub-dot should be a scoped session: its own context, tools, instructions, and config. The main dot delegates. The sub-dot does not inherit the whole brain. Think of it like a plug-in made specifically for dots. 4. A heartbeat. Default cadence, overridable per task. "Watch this" should mean every N minutes, not whenever proactive research feels like it. Teams of dots were the promise. Sub-dots are how you get there without waiting on a second primary. Now that would be a meaningful improvement or upgrade. But since none of that happened, how about a bank reset? 🤷‍♂️🤪🫣 @thsottiaux @ChatGPT @OpenAI @OpenAIDevs nitter.gq/roofixai/status/210860…
Replying to @thsottiaux
@thsottiaux What happened to yesterday’s reset? 🤔 It’s not the end of the world, but the only real update was UltraFast for $500/mth users. Basically, we $200/mth users are no longer @OpenAI’s most valuable users. FYI—that 50% less usage really stings. I figured the daily reset was just a way to keep us $200 users happy for another month or so. Either way, please stay true to your word on these resets this one time. I’d love to power through this month’s work with 6.1-Sol and Astra without worrying about daily usage. I just used my last banked reset. Every time I use one, you push a broad auto-reset. If most resets will be auto-resets, give us a daily window for the decision/announcement so I can plan my day without the extra stress. lol 🙏 Thanks 👍 @MatthewBerman @sama @CodexReleases
1
1
1
2
132
RooFix AI retweeted
Replying to @thsottiaux
@thsottiaux I’m just saying…
1
1
1
17
$200/month Codex Pro 20x is giving free/plus tier vibes this week.
5
112
RooFix AI retweeted
I think I can officially say: we are so back
2,038
489
150
19,311
1,430,638
Just tapped my second $$ tokens auto reload. 🫣 @thsottiaux - is it time to circle back to the circling back of your initial circle back? 🤷‍♂️
🤖 Made with AI
Replying to @viewsfrom02108
Let's circle back soon
1
4
139
Hey @vercel - I really need you to fix your deployment issue as quickly as possible. All deployment attempts are currently stuck in the "initializing" state. @vercel_status @vercel_dev @vercel_support
🤖 Made with AI
14
12
206
RooFix AI retweeted
GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We’re expanding preview access globally now.
2,414
6,573
4,080
46,773
9,979,778
RooFix AI retweeted
I didn't expect this tweet to be controversial. But here we are. Okay so look at the most popular apps you use every day. Now think about what they looked like two years ago vs today. They're all converging on the same three panels. An explorer on the left. A chat in the middle. A preview on the right. Look at the new apps being built. Same shape. That's Slack. Now think about how your team actually works. Work starts and happens in Slack. You only tab away for two things: context and ui. Context, because the data you need lives somewhere else. UI, because some things need buttons. Different buttons. An agent collapses both into one app. Context, it's already good at, it goes and gets whatever you need. UI, it's getting there. Soon it'll build the buttons right where you work. Whoever nails this shape owns where work happens next. The 2030 Microsoft Office.
"Slack is close to being the perfect UI for most apps." nitter.gq/shadcn/status/19833882…
115
64
25
1,810
299,248
If you care about the American dream and entrepreneurship in the age of AI…you need to watch this 👇 @MatthewBerman
5
6
106