Understudy
Open-source desktop agent that learns from demonstrations
What it does
Teach a desktop agent by doing a task once. Understudy learns the workflow, turns it into a reusable skill, and can later execute it for a new input across GUI apps, browsers, shell tools, files, and messaging. It’s open source, local-first, and designed for real cross-app work instead of brittle macro playback. Current state: macOS today, MIT licensed, with teach-by-demonstration working and route optimization still early.
Does a similar job
all alternatives →- UTUnderstudy – Teach a desktop agent by demonstrating a task onceMar 2026 · github.com · ▲120
I built Understudy because a lot of real work still spans native desktop apps, browser tabs, terminals, and chat tools. Most current agents live in only one of those surfaces. Understudy is a local-first desktop agent runtime that can operate GUI apps, browsers, shell tools, files, and messaging in one session. The part I'm most interested in feedback on is teach-by-demonstration: you do a task once, the agent records screen video + semantic events, extracts the intent rather than coordinates, and turns it into a reusable skill. Demo video:…
- OAOpenWork – An open-source alternative to Claude CoworkJan 2026 · github.com · ▲231
hi hn, i built openwork, an open-source, local-first system inspired by claude cowork. it’s a native desktop app that runs on top of opencode (opencode.ai). it’s basically an alternative gui for opencode, which (at least until now) has been more focused on technical folks. the original seed for openwork was simple: i have a home server, and i wanted my wife and i to be able to run privileged workflows. things like controlling home assistant, or deploying custom web apps (e.g. our customs recipe app recipes.benjaminshafii.com), legal torrents, without living in a terminal. our initial setup…
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 · 18d 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 · 28d ago · cactuscompute.com
Launched alongside, March 2026
the whole month →
Switch from ChatGPT to Claude with import memory feature
AI · Mar 2026 · claude.com
