Lobster Roll

I’m ready to retire. In my younger days, I remember a few pivotal moments for me as a young nerd. Active Server Pages. COM components. VB6. I know these are laughable today but back then it was the greatest thing in the world to be able to call server-side commands. It kept me up nights trying to ab...

Hi HN, I forked chromium and built agent-browser-protocol (ABP) after noticing that most browser-agent failures aren’t really about the model misunderstanding the page. Instead, the problem is that the model is reasoning from a stale state.<p>ABP is designed to keep the acting agent synchronized wit...

Hi HN, we&#x27;re Sanchit and Shubham (YC W26). We built a fast inference engine for Apple Silicon. LLMs, speech-to-text, text-to-speech – MetalRT beats llama.cpp, Apple&#x27;s MLX, Ollama, and sherpa-onnx on every modality we tested. Custom Metal shaders, no framework overhead.<p>Also, we&#x27;ve o...

I found that duplicating a specific block of 7 middle layers in Qwen2-72B, without modifying any weights, improved performance across all Open LLM Leaderboard benchmarks and took #1. As of 2026, the top 4 models on that leaderboard are still descendants.<p>The weird finding: single-layer duplication...

Agent Kanban has 4 main features:<p>GitOps &amp; team friendly kanban board integration inside VS Code Structured plan &#x2F; todo &#x2F; implement via @kanban commands Leverages your existing agent harness rather than trying to bundle a built in one .md task format provides a permanent (editable) s...

See also the [release announcement post](https://determinate.systems/blog/determinate-nix-future).