OpenDots: an open-source replica of the OpenAI Dots workspace
OpenDots, an MIT-licensed open-source project, reached 2.8K stars in five days by letting anyone run a local team of specialist bots with their own computers, approval flow and voice calls.
THIS IS FREAKING INSANE! OpenAI Dots - 2.8K stars repo in 5 days! free treasure for everyone who wants their own FREE team of OpenAI-style Dots i found the repo that gives you an open-source Dots workspace: specialist agents, each with its own computer, that you can text, call, or mention in Slack. it went from 1.3K to 2.8K stars in two days. it's called OpenDots, it's MIT licensed, and it runs on your own machine. here's the breakdown: → agents with jobs, not one chatbot. - you give each Dot a name, a role, instructions and its own list of allowed tools. one can be your researcher, another your writer. they get separate conversations, and you control exactly what each one is allowed to touch. → every Dot gets its own computer. - each one can have its own browser, files and terminal, and the browser profile and files survive restarts. browser, file and shell permissions are set per Dot. there's a live computer panel where you watch it work, and you can take over the browser at any moment. → nothing gets saved without you. this is the part i like most. ask a Dot for a draft and the chat pauses on a card with two buttons: approve & save, or decline. only after you approve does it become a page in your workspace. the agent waits for your call before it keeps going. → a real workspace, not a chat log. - approved work lands in "Spaces": a searchable library of pages with a visual editor, slash commands, nested subpages and autosave. every page has its own conversation, so you can ask a Dot about a document right next to it. → you can literally call it. - there's a voice mode with live captions, mute and a call timer. the clever part is that speech and work run separately: you keep talking while a second agent handles the longer task in the background. → it lives where you already work. - mention a Dot in Slack and it continues the work in the thread, with an allowlist that controls which workspace and users can reach it. → it can work on a schedule and learn. - scheduled tasks run in the background with pause and retry, and "Automatic Learning" turns patterns from your - conversations into reusable skills. → setup is four commands. - clone it, npm ci, copy the env file, npm run dev. you can create Spaces and configure your Dots before connecting a single service. i track every repo in this Dots wave so you don't have to. stay tuned to catch the next one before it hits 10K, and show this for the friend who's about to pay for something they could self-host. so here's the real question: if anyone can clone a team of always-on agents in four commands, what exactly is OpenAI charging for?
This post quotes another post: x.com/teka1900/status/2105602481506156562