I started building a React file uploader because I needed one.
Halfway through, I found myself asking a different question:
Do we still need open source libraries when AI can generate components?
That question completely changed how I approached the project.
Instead of just building another uploader, I experimented with what an OSS library should look like in the era of AI-assisted development—headless architecture, documentation for both humans and AI assistants, llms.txt, and Context7 support.
I wrote about the journey here:
🔗 https://t.co/7gUtqCF31C
I'd love to hear how other OSS maintainers are thinking about this shift.
Introducing https://t.co/8MeR16MRQn, a tool to measure how well agents can read your site. Backed by @oradotai's research, you can run:
▪︎ Audits with 100+ checks
▪︎ Visualizations of agents using your site
▪︎ One-click prompts to fix problems
▪︎ A CLI for agents
We're making Git hosting more reliable, performant, and scalable.
This post traces 20 years of Git infrastructure and explains how that history led us to design and operate our Git storage, Origin, as if it were a database.
https://t.co/UW7jHuItSX
Good read! Today, when I was thinking why @Lovable chose @tan_stack over of @nextjs while building an app and there were no concrete answers available.
Now, got the answer in surprise. 😃
@antonosika Value proposition is not an outlier. Talking to people will break the limbo in our mind about the product we are building and what it really means for the user.
@zenorocha I saw this pattern on many sessions when I inspect what agents are thinking. They want to take authority when building and invoke tools unless asked.
But “it works” moment change when we ask them with more to build and finally end up with non maintainable code
@svpino Folks, who codes more than a decade still follows/applies the same principle they learned over the years. Just the execution is delegated to a thinking agent.
In my experience, the cost of building software hasn't decreased at all.
People are still working 40 hours every week. They are still putting in maximum effort to build whatever they are building.
Are we shipping more and better software? Likely yes.
But we are also paying a ton of money in tokens, and I don't really see companies firing half their development teams to save money.
if I had to start a new startup from scratch, I'd pick a very boring infra problem.
boring problems are owned by lazy giants.
and lazy giants have legacy SDKs and confusing docs.
in an agent's eyes, that's the whole product.