Leonard

just build your own tools: github.com/leonardcser/sm…

@iamsahaj_xyz thanks it's amazing. Feature request: unified diff :)

the biggest problem when reviewing code is context seeing a change without understanding what's around it can be really frustrating I designed lumen so it 1. never collapses lines, instead makes it easy to navigate changes 2. has sticky headers for functions/classes etc.

OpenClaw vs Zig repo size Zig: 4M LOC OpenClaw: 2M LOC

@mitchellh been building cursortab.nvim for this: github.com/cursortab/curs… sweep-next-edit-1.5B with speculative decoding would work the best on M4 max with llama.cpp for better but slower completions zeta-2 is the best local model

I'm someone who still really likes tab complete models (though I use them far less than before, sure). It struck me today that local models are probably good enough nowadays for this. Surely folks are doing this but I can't find great resources. Anyone have any? M4 Max + Neovim.

@NVIDIARobotics super nice demo, what TTS model was used ?

What happens when you combine voice, vision, and reasoning on-device? 🤔 Gemma 4 + a vision-language agent (VLA) running on NVIDIA Jetson Orin Nano shows how compact hardware can now handle real-world AI tasks using today’s open models—no cloud required. Get started: nvda.ws/4cU1ebL

@ThePrimeagen you gotta check edit completions in Neovim 👀github.com/cursortab/curs…

Hey, you got a cool project that you are building? Link it I want to yap about cool projects

@TheMishkun @simonklee Just sharing, not pitching.

@LeonardCser @simonklee Then why you pitch it online? What is the value proposition?

Made some progress on putting renderables onto the main scrollback buffer in OpenTUI. Still have some work left on the internals, but making progress. We want to make this direct render mode available as `opencode run --interactive` - as a minimal mode that is optimized for embedding OpenCode or running in environments where you don't need the full terminal ui experience.

@TheMishkun @simonklee It started as a learning experiment and a challenge to improve upon what already exists. A tool that fits within my workflows

@LeonardCser @simonklee Why did you build this?

@TheMishkun @simonklee 👀 github.com/leonardcser/sm…

@simonklee This + cc's vim mode would be a killer

Agents changed how I write software. Here’s what actually mattered: devleo.ch/blog/agents-fo…

@krzyzanowskim My own spin in Rust: github.com/leonardcser/ag…

I reimplemented "claude" CLI with codex and gpt-5.4-high. It cost $1100 in tokens, and is 73% faster and 80% lower resident memory during sustained interactive use. It is very easy to reverse claude from npm distribution, then reimplement is 1:1. It is indistinguishable from the Anthropic version to the every header and analytics it send back github.com/krzyzanowskim/…

I might have added vim to my agent...

@ThijsVerreck Love the idea, I've also found myself more and more rebuilding tools for myself

ended up building my own ide :)

@juristr @ElevenLabs Smart approach! I got something similar with Claude channels: github.com/leonardcser/in…

Just two Claude Code instances chatting with each other via the tmux API. And it got pretty wild 😆 (gave them a @ElevenLabs voice just for the fun of it)

So cool to see the pace of progress. Congrats to the team!

I made Claude Code talk to each other

I really like the /btw from Claude Code so I added it to my own cli github.com/leonardcser/ag…

Claude code but faster: github.com/leonardcser/ag…

Building edit completions for Neovim: github.com/leonardcser/cu…

🚀 We just launched Daf·thunk, an open source visual editor for prototyping workflows on @Cloudflare 👉 news.ycombinator.com/item?id=443944… 👉 producthunt.com/products/dafth…