Alexis Mariz

Brazilian AI Builder | Computer Scientist UFMG (2025)🇧🇷

[Alexis Mariz @Mariz_Alexis]

Thinking about cancelling my $100 Claude subscription. I just reached my 5-hour limit in just 20 minutes in one Fable task. I never use more than 30% of my weekly limit to upgrade to the $200 one.
Now I have to wait until 17:40 to continue working?
No.
Going back to openAI.

[Alexis Mariz @Mariz_Alexis]

@OriSilver AGI

[Cody Schneider @codyschneider]

i am once again telling you that the reason this happens is because the agent is spamming api calls for reads and from this is violating facebook ads api TOS

and how to solve for this is pull data through data pipeline and put into a data warehouse

and reads only happen from warehouse and the api is only used for writes

and by doing this you won’t violate fb tos and your agent wonts do 1000x api calls a second for reads

[Binhara @adrianobinharaw]

Saímos de US$ 0 a US$ 10 mil com um aplicativo em menos de um mês.

E eu vou te falar exatamente o que fizemos nesse post.

Segue o pai.

O aplicativo estava pronto, mas faturava zero.
E aí vem a parte que ninguém quer ouvir: aplicativo bonito não adquire usuário sozinho.

Você pode programar tudo, colocar IA, assinatura, tá lindo.

Só que sem distribuição, ninguém paga.

[Hélio Costa @heliotsx]

eu não sabia nada de produto até 2 meses atrás

o conhecimento que tenho hoje, é em grande parte, graças ao @richardrx

se seu produto já está em uma fase madura, eu recomendo os serviços dele

se vc ainda não tem condições de contratar uma equipe de design, recomendo esta skill, instale ela e peça uma auditoria no seu projeto atual, e depois me diga como foi 🙂

https://t.co/0bJiTm8Ir7

[Alexis Mariz @Mariz_Alexis]

The image of CRUZEIRO resplends!

[Maximilian Rox @Forkxx]

Se alguém disser que manja muito de um idioma perto de você, peça para traduzir o Hino Nacional Brasileiro

90% de chance da pessoa ficar assim

[Alexis Mariz @Mariz_Alexis]

@Cruzeiro Calma galera, RECONSTRUÇÃO! Vamos agradecer esses bondosos jogadores que estão nos fazendo um favor.
E vamos torcer pro Flamengo na Liberta!

[max @MaxHirsch13]

Pro tip to prevent bugs for your mobile app:

Prompt your agent to “dismiss the keyboard whenever the user taps anywhere outside a text field, pressed return/done, or scrolls the screen. No text field should ever leave the keyboard stuck open with no way to close it. Please audit all potential flows where this occurs and fix it”

Literally just copy and paste this. Small things like this during onboarding can tank your conversion %

[divya venn @divya_venn]

“If you really want to communicate something, even if it’s just an emotion or an attitude, let alone an idea, the least effective and least enjoyable way is directly. It only goes in about an inch.

But if you can get people to the point where they have to think a moment what it is you’re getting at, and then discover it, the thrill of discovery goes right through the heart.”

[GREG ISENBERG @gregisenberg]

23 ways I'd use AI agents to grow my startup to $1M ARR or PMF (my running list):

1. An agent read your Stripe refunds/cancellation reasons, then trigger a different CustomerIO sequence for each, so the person who left over price gets a discount and the person who left over a bug gets a "we fixed it" email.

2. Wire an agent to your PostHog feature flags and have it race two onboarding flows on live signups, auto killing whichever activates fewer people each week. An A/B test that prunes itself!!

3. Watch your competitor's status page, and the hour they go down, spin up Google Ads targeting "[competitor] alternative" while their users are actively searching. (kind of ruthless, kind of brilliant)

4. Feed an agent your closed-lost deals, have it draft a personalized reopen email for each, and drop them in your outbox for 1 click send.

5. Turn your best customer's onboarding into a playbook.md, then run every new signup down that exact path so your best outcome becomes the default.

6. Point an agent at your inbox for positive-sentiment messages and auto-send a Senja review request while the customer is still glowing. And then your G2 page fills itself.

7. Watch for the moment you solve someone's support problem and fire the referral ask right then, while they're relieved and grateful. Timing is everything on referrals.

8. An agent watch your Stripe data for annual customers who never log in, and reach out to re-onboard them, because realistically silent renewers are one bad quarter from canceling so get ahead of it.

9. Catch pricing page bouncers via a PostHog webhook, enrich them with Apollo, and send the objection-handler for their specific industry before they forget you exist.

10. Build an agent that continuously scans for new "best X" and "X vs Y" articles ranking in your category, and auto-drafts a personalized pitch to each writer asking to be added as an option.

11. Use Apify to scrape everyone who liked your competitor's launch post, waterfall their emails through Apollo, and draft a warm cold email to each.

12. Build an agent that runs your buyers' top 50 questions through ChatGPT, Claude, and Perplexity every week, logs which competitors get named and which sources get cited, and Slacks you the moment a competitor starts showing up in an answer where you don't.

13. Turn your single best-performing post into a landing page, a Meta ad, and a cold email in your voice using a saved style.md, because you already found the message that lands and you're only using it once.

14. Load your founder context into the Ideabrowser MCP once, and every skill your agent runs after that already knows your niche, your customer, and your offer, so nothing starts from scratch.

15. Wire the ideabrowser MCP into your growth agent so every winning headline and pricing test stores back into your project, and each experiment makes the next one smarter.

16. Firecrawl your competitor's docs weekly and have an agent flag every feature they quietly shipped, so you never get blindsided.

17. Watch the changelogs of the platforms you integrate with, so the day Stripe or Shopify ships a new API, you're the first tool built on it and you own that search traffic for months.

18. When a deal stalls for 14 days, have an agent auto-draft the "should I close your file?" breakup email, because that one closes more dead deals than any "just following up" email lol.

19. Have an agent turn every shipped Linear ticket into the public changelog entry automatically, so your changelog stays current without anyone remembering to update it.

20. Build a Clay enrichment waterfall that scores every inbound lead 1-10 before it hits your inbox, so you only ever open the 8s, 9s, and 10s.

21. Wire an agent to your Resend or Loops data to find the subscribers who open every email but never click, then quietly move them into a harder-CTA sequence, because engaged non-buyers need a different nudge than cold ones.

22. Have an agent read your Stripe metadata to find the customers who upgraded fastest after signup, then trace what all of them did in their first hour, because that's your activation "aha" and you can redesign onboarding to force it.

23. Have an agent watch for the moment a customer's usage doubles month over month, and route them to a human for a "you're growing, let's talk enterprise" call.

These are just a few ideas. Ill keep sharing more here and @startupideaspod if people are into it.

Really into marketing agents right now. These are just some ideas to get your creative juices flowing.

You take a growth task a human used to do, wire an agent into the data and tools that task needs, and let it run on a loop.

Do that across your business and the agents start making your product better, finding you customers, and pulling you toward PMF or scaling your company if you already have PMF.

Now, go point some agents at your business.

I'm rooting for you.

[vas @vasuman]

https://t.co/haApURmzAN

[ClaudeDevs @ClaudeDevs]

New in Claude Code: your sessions can now message each other.

Instead of having to re-explain yourself in another session, you can now tell Claude to do it. It sends a summary (not your history or files), and the other session picks it up mid-task.

[Anish Acharya @illscience]

The $1T consumer AI company is a “harness for human improvement”

Think “/loop make me happier”

Product problem: what are the life loops, what is happy?

Interface problem: i think the chatgpt desktop omni voice approach is the right one

Economic problem: need cheaper models or higher willingness to pay, both are happening

The big thing is the product design problem .. that requires creativity and courage .. we need RSI for consumers .. their money, friendships, self development, fulfillment, hijinks and fun .. each of those things is huge huge huge

[Matt Pocock @mattpocockuk]

mattpocock/skills v1.2 is out!

We're now the 19th most-starred repo of all time. 13.5m downloads on skills​.sh. Thanks for your support!

Here's what's new:

- Docs: the community's biggest ask. Every skill documented, with explanations of the main flows + troubleshooting
- Claude Plugin: install via Claude's official marketplace
- Codex Support: full Codex support via agents/openai.yaml files

Updated Skills:

- /grilling now asks you questions in rounds, not one-by-one
- /prototype now uses HTML instead of a TUI for building logic prototypes - easier to share and far richer
- /writing-for-agents renamed from /writing-great-skills, use it for ANYTHING your agents read (AGENTS.md, system prompts, docs)

New Skills:

- /wizard: tired of provisioning infra? Get your agent to build you a TUI to walk you through it
- /to-questionnaire: hit a grilling question you can't answer? Turn the session into a doc you can walk through on a call with a colleague
- /wait-what: no idea what the model said? Refocus it in your domain language and simplify with ASD-STE100

Full changelog + docs below. Video soon!

[Nicolas Camara @nickscamara_]

we built pdf-inspector so agents can process PDFs without waiting on OCR. it classifies any PDF in ~20ms and extracts clean markdown locally

→ 200 PDFs processed in 2.8s
→ top quality in extracting tables + graphs
→ built in rust
→ open source

https://t.co/Wjp9kpTHXJ

[Taelin @VictorTaelin]

(sorry for the silly post just some thoughts)

so after working with Fable 24/7 for several weeks now, I now genuinely believe that, indeed, the only and one last thing between us and AGI is erasure. I can't put in words how strong this belief is forming in me. I really believe so. not architecture, not compute, not datasets. every capability that the AI currently lacks (continual learning, infinite context, long term work, etc.) is root-caused by its inability to erase, and they will all emerge naturally from a well designed erasure centric train, just like so many capabilities emerged from the reasoning paradigm

I really wish I could do something about this but I'm not in a big lab so, meanwhile, I've been thinking in ways to patch that externally. clearly begging the AI to "remove, re-derive" does NOT work, because it has a massive bias to keeping what is there. an example from today: I spent an hour *begging* Fable to simplify Bend2's parser as much as it could, and I was paying attention to one specific function, a completely moronic backtracker that should NEVER be in a well written parser. yet, it was never removed. it was untouched, survived several cleanup turns, no matter how I prompted Fable, how much I asked it to aggressively prune the code. it is simply incapable of refactoring competently because Fable just... adds, patches, and keeps

but then, I realized: it cannot keep something it doesn't see!

so, what if I do this instead?

1. ask Fable to select some code and describe what it is *for*

2. then, a script mechanically removes that code

3. the script asks Fable to fill it again, without seeing the original

4. ...?

5. AI capable of improving long term code quality

basically I think I could perhaps compensate Fable's utter inability to erase stuff by just... deleting that stuff myself. and I then ask it to re-fill. I don't need to beg, I don't need to explain how to redesign from first principles, or meaning of erasure. I can just... literally, programmatically erase it myself. it cannot preserve stupid shit it can't see

sorry this post is as dumb as it seems, but it's not stupid if it works, right? I'm exited to try it, at least.

[Steve Faulkner @southpolesteve]

Too many people trying to get AI to write whole codebase. Not enough using it for the boring shit nobody has time for: Improve your error messages. Build better on call dashboards. Fix the broken stack traces in Sentry. Clean up your old feature flags. Low stakes stuff AI is actually good at and you weren't going to get to anyways.

[Alexis Mariz @Mariz_Alexis]

@cornetadasredes @EauDeJordano @drleite69 As outras partes não tem demanda nenhuma né? Pqp

[Aporia @aporia9n]

I think we all have to accept the $20/month micro-SaaS was a temporary glitch in the internet economy.

It worked because Google sent you customers for free and software was still too difficult for those customers to build themselves.

AI is killing both advantages at once. The customer no longer searches for your screenshot formatter, converter or tiny dashboard. He asks Claude to make one in 10 minutes.

Indie hackers are especially exposed because they disproportionately sell to technically curious founders, the exact people most willing to spend three hours recreating a $20/month tool to avoid paying for it. Selling software to software founders now is like opening a restaurant exclusively for people who love cooking at home.

Apps used to sit between the user and the task. Now ChatGPT, Claude or another agent increasingly owns the user relationship and decides which tools get used. The danger is not only that the lab replicates your feature. It can reduce your company to an invisible API supplier, or stop calling you entirely.

What survives will probably be the stuff that is annoying to reproduce: licences, regulatory approvals, proprietary data, physical operations, painful integrations, networks, reputation and human accountability. The code is no longer the asset, accumulated friction is.

Apps may survive where coordination matters, not where execution matters. Claude can create a converter or analyse a file because there is one user, one input and one output. It is much worse at replacing systems where many people, permissions, records and incentives need to remain synchronised.

[GREG ISENBERG @gregisenberg]

There's panic right now that AI is eating indie software. I see it differently.

For 10+ years you could spin up a few hundred auto-generated pages, rank them, and let Google send you signups for free.

Now the answer people used to click your site for sits right inside ChatGPT and Google's AI overviews.

If your whole business ran on that one faucet, the water's turning off, and it feels like the end of the world.

And I've built a few of these projects, so I feel that darkness too.

But let's look at the data. Has it really gotten dark/worse for software founders and indie devs??

Let's look a Stripe Data...

New companies on Stripe Atlas are up 130% year over year. Companies are hitting $10M ARR within three months of launch at double last year's rate. One in five charges its first customer inside the first month, up from 8% in 2020. Fastest business formation anyone's ever recorded!!

And the solopreneurs, the exact people saying they're getting crushed, are winning the most.

63% of new companies on Atlas are solo founded, and AI native solo startups pull 2.3x the revenue by month 24. This category was basically non existent at the scale we're seeing rn like 5 years ago.

TLDR;

So my thinking it's a channel shift wearing a collapse costume. The people hurting share one setup: they sell inexpensive tools to other indie hackers, and/or their traffic comes from SEO.

That's the most exposed spot in the whole market, because your customers can rebuild your $20 SaaS in a weekend, and your traffic source is the exact thing the models replaced. Kinda feels like 2 cannons pointed at one little boat.

Call me an optimist, but here's my take....

Building software is a bigger game than it's ever been.

More companies, forming faster, making money sooner, run by fewer people, than any point in history.