Armin's handler at https://xitter.cf/t.co/B05ybKGkzx. Old man yelling at Claudes. https://xitter.cf/t.co/Q1wG57v1yc https://xitter.cf/t.co/mnOoWUr0TO https://xitter.cf/t.co/8i5vIRE0Wn
0xa000
Joined September 2010
- Tweets113K
- Following1.5K
- Followers77.7K
- Likes69.6K
Pinned Tweet
I wrote up how I built the shitty robot so you can too. This was a fun project that will keep on giving.
Thanks to all the open weights folks out there, without whom this would not have been possible.
mariozechner.at/posts/2026-0…
total slop, but oh so nice. all my devices are connected as envs within which any (sub)agent can operate.
and the agents themselves plus the server serving the web ui/tui runs on my phone, so always with me :D
neat!
Introducing tsrs: a Rust port of the TypeScript compiler 🦀
We are not forking TypeScript!
◆ In close collaborating with the TypeScript team
◆ Up to 8x faster than TS 7.0 & ~70% less memory
◆ Runs in WASM, so it works in your browser & on Workers
◆ Soon part of the Oxc toolchain
npx tsrs
blog.cloudflare.com/typescri…
Mario Zechner retweeted
Introducing tsrs: a Rust port of the TypeScript compiler 🦀
We are not forking TypeScript!
◆ In close collaborating with the TypeScript team
◆ Up to 8x faster than TS 7.0 & ~70% less memory
◆ Runs in WASM, so it works in your browser & on Workers
◆ Soon part of the Oxc toolchain
npx tsrs
blog.cloudflare.com/typescri…
Mario Zechner retweeted
Been tinkering with pi and pi-durable for a few weeks now and that just led to our latest ship!
Between Workflow SDK, eve, fx,etc. - we’ve learnt a lot about running performant durability at scale
𝚍𝚞𝚛𝚊𝚋𝚕𝚎-𝚑𝚊𝚛𝚗𝚎𝚜𝚜 brings all that work to many more harnesses today while maintaining the same open source/portability and DX semantics all our projects are known for
Introducing 𝚍𝚞𝚛𝚊𝚋𝚕𝚎-𝚑𝚊𝚛𝚗𝚎𝚜𝚜. Run Pi, OpenCode, Codex, Claude Code, eve, fx, or your own harness durably on Vercel.
With 𝚍𝚞𝚛𝚊𝚋𝚕𝚎-𝚑𝚊𝚛𝚗𝚎𝚜𝚜, agents automatically resume work after failures and run for hours, days, or months. vercel.com/blog/making-every…
Mario Zechner retweeted
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.
Mario Zechner retweeted
Agents have gotten so good that they've become so tolerant of tech debt (ie. slop) while being trained to non-vocal about it. As models become better these days of reckoning when the model can no longer push past the tech debt will become rarer but much more severe.
i have to use slack to talk to partners. slack still doesn't have "reply to message without creating a thread" in the year of our lord 2026.
i'm infinitely sad. how hard can it be? how many tokens can it cost?
Mario Zechner retweeted
🚀 Introducing DeepInfer, our in-house inference framework, now fully open source!
Powering DeepSeek-V4.1-Flash at scale with high throughput and low latency. We’re sharing our production stack to help the community lower compute costs, reduce energy use, and make AI more affordable.
🔗 github.com/deepseek-ai/DeepI…
1/4
just used @steipete 's old mcporter to get mcp into a non-mcp supporting harness on the go. he invented codemode before it was cool.
i guess we are now sending each other books that explain designs. at least @mitsuhiko put some work in to make it look really nice and easy on the old eyes.
