ViewSpec
Universal UI compiler for AI agents
What it does
I built ViewSpec because my AI agent kept struggling with navigating the web. DOM was never designed for non-human users. ViewSpec is a universal intermediate representation for UIs. You describe what you want — metrics to show, actions to enable, layout constraints — and it compiles into a polished, accessible interface. Think Tailwind for agentic apps. The key idea: instead of asking agents to reverse-engineer human UIs, give them a stable semantic layer that's legible by construction.
Does a similar job
all alternatives →
Webvizio for AI Coding Agents2025 · webvizio.com · ▲706Resolve dev tasks & bugs faster with Webvizio MCP
- AVA Visual IDE for React2022 · dev.aspect.app · ▲479
I made this because building UIs in a lexical medium like code is super annoying. I have to pre-render what I’m making in my head, and then jump between the browser and IDE to test. I was inspired by the developer console in chrome and safari since I end up editing css there because it’s ironically more convenient. Hope it’s useful!
- DODAC – open-source dashboard as code tool for agents and humansApr 2026 · github.com · ▲119
Hi all, this is Burak. When agents became a reality one of the first things I wanted to do was to automate building dashboards. The first, and the most obvious, wall that I ran into was that a lot of the tools were just driven by UI. This meant that without the agents handling browser UIs and whatnot, it wasn't possible to have the agents do that. In addition, it would be impossible to review any of the changes the agent would make. The first instinct there is to get your agent to build a React app for the dashboard. This works beautifully for the happy path, but I quickly ran into other…
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 19d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, May 2026
the whole month →
Parallel agents, diff reviewer, and multi-model comparisons
Dev tools · May 2026 · kilo.ai
- NW
Hey HN, Henry here from Cactus. We open-sourced Needle, a 26M parameter function-calling (tool use) model. It runs at 6000 tok/s prefill and 1200 tok/s decode on consumer devices. We were always frustrated by the little effort made towards building agentic models that run on budget phones, so we conducted investigations that led to an observation: agentic experiences are built upon tool calling, and massive models are overkill for it. Tool calling is fundamentally retrieval-and-assembly (match query to tool name, extract argument values, emit JSON), not reasoning. Cross-attention…
Life & fun · May 2026 · github.com
- FM
Dev tools · May 2026 · github.com