Introducing pyRPC
tRPC-style type safety for Python backends. Define your procedures in Python, call them from TypeScript with full inference. Auto complete, type inference, No OpenAPI.
Change a Python type -> TypeScript catches it instantly.
https://t.co/KEBTXeN7nH
We all have ideas. Ideas are immortal. They last forever.
What doesn’t last forever is inspiration. Inspiration is like fresh fruit or milk: It has an expiration date.
If you want to do something, you’ve got to do it now. You can’t put it on a shelf and wait two months to get around to it. You can’t just say you’ll do it later. Later, you won’t be pumped up about it anymore.
If you’re inspired on a Friday, swear off the weekend and dive into the project. When you’re high on inspiration, you can get two weeks of work done in twenty-four hours. I
Inspiration is a time machine in that way. Inspiration is a magical thing, a productivity multiplier, a motivator. But it won’t wait for you. Inspiration is a now thing. If it grabs you, grab it right back and put it to work.
Inspiration is perishable.
@orcdev https://t.co/GtNhyDjo1S - an open-source rpc library for e2e type-saftey between Python backends and TypeScript frontends. (basically tRPC but for Python backends)
Support for FastAPI, Flask, Django, React, Next.js, Vue, Svelte adapters. easier to set up with simple syntax.
@orcdev https://t.co/GtNhyDjo1S - an open-source rpc library for e2e type-saftey between Python backends and TypeScript frontends. (basically tRPC but for Python backends)
Support for FastAPI, Flask, Django, React, Next.js, Vue, Svelte adapters. easier to set up with simple syntax.
Video Podcast is live.
Turn any doc, link, or idea into a two-host video show with studio scenes, multi-cam cuts, and B-roll. Ready in minutes.
Everyone else stops at audio. This is a show you can publish.
Try it: https://t.co/mwV67pFCDT
@cursor_ai 6/
We didn't get to build everything we had planned, but we delivered a working prototype and finished in the Top 15.
Huge thanks to the Ethiopian Cursor Community, everyone who made the event possible, and my teammates for making it memorable.
Looking forward to the next one.
24 hours building at the largest AI hackathon in Ethiopia.
This past weekend at the @cursor_ai Hackathon Ethiopia 2026, our team explored one question: How do you know you're paying a fair price for something in Ethiopia?
#CursorAI#Hackathon#AI#Ethiopia
read more...
@cursor_ai 5/
Whether you're looking for a phone in Addis Ababa or furniture in Hawassa, the goal is the same: better understand the local market and make better-informed decisions.
@cursor_ai 4/
It helps people:
- Estimate a fair local market price using information from existing marketplaces,
- Compare nearby options
- See community insights about product quality
@cursor_ai 3/
Instead of building another marketplace, we explored how AI could make the existing market easier to navigate. That became Zabiya - a location-aware shopping assistant built for the Ethiopian market.
@cursor_ai 2/
Today, much of Ethiopia's commerce happens across Telegram channels, Facebook Marketplace, TikTok, and individual storefronts. Products are easy to find, but understanding the market around you isn't.
I've always been team @nextjs, but the caching footprint is starting to feel heavy.
Had to clear disk space and noticed multiple projects with multi-GB .next folders.
Any solid patterns for keeping .next sizes reasonable without killing the DX?
My advice whenever someone asks me for a recommendation on their YC application.
Politely: No thank you.
YC hates it when founders seem to care more about getting into YC than they do about their customers. The way to get in is by making progress on your company not being obsessive about YC.
For anyone that does something silly like 'shooting their shot' by sitting outside the YC offices or tagging a bunch of YC Partners on an X post - that's not shooting your shot.
Shooting your shot is:
Sitting outside the target customer's office until they try your product.
Flying to offer a key customer lunch just so you can hear what's frustrating them
Writing amazingly targeted, helpful content that shows your customers that you're the one they should go to for more help.
Relentlessly finding new customers and making them happy by fixing every problem they have with your worst-it-will-ever-be early product.
Learning something each week from your users so that today's vision of the product changes dramatically each week.
Do all this and you'll grow.
Tell the stories of doing all this in your YC application. That's shooting your shot.
This is yet another reason it's good for young founders, especially, to start by building something they themselves want. If you're the intended user, you're not going to be satisfied with vague. (Nor is any user, but at least this way you're forced to see it.)
An initial startup idea can't usually be both grand and precise. In practice they're usually either grand and vague or precise and small. Precise and small is better. You know who your initial users are, and you expand outward. With grand and vague you can't even get started.
@JasperLind87783 This AGENTS.md is deliberately short & high-level (just repo-specific rules). Even smaller-context models handle it fine via Herdr. No stability issues so far when switching between Claude/Codex/OpenCode.
This L8 engineers' agentic setup is amazing, but it only runs on macOS.
So I ported it to windows.
nix + home-manager + neovim + herdr + one shared AGENTS.md across Claude/codex/opencode - all running on WSL2, tested end to end tonight.
https://t.co/9rk0YZT2uG
many people asked me to make a video about my complete agentic engineering workflow
excited to share it's finally here!!!
it took me about 20 hours in total to record this 45 minutes of walkthrough - it covers everything i do to ship production quality code at an average 40+ PRs/day velocity
hope this can be a useful reference to everyone exploring good ways to use AI. and would appreciate a reshare with anyone you think might benefit from this!
enjoy! https://t.co/oA0UCrBvqo