A very detailed explanation and setup video for @MulticaAI. If you're using Claude Code, Codex, or the @NousResearch Hermes Agent, you can follow this video to set up your workspace. It's very easy. You don't need to do this manually. Just tell your agent: "Help me set up Multica for xxx"

@DimAgentai is officially supported by @MulticaAI . Now you can use dim and dim model provider in multica. We set an automatic goal to let dimagent run in one week to integrate itself into multica. Pretty much thanks to @jiayuan_jy and Multica community.

I wrote a tutorial on how to use @MulticaAI for scientific research. Multica is an open source agent-first project management platform that gives you an army of agents you can delegate tasks to. It's like having your own team of scientists and researchers. Read the tutorial to get started:

Handing work between agents without losing everything in the process is crucial. We connected our Scientific Agent Skills library to Multica, an open-source agent task board. Ran a real ALS (Lou Gehrig's disease) trial analysis end to end: 478 records pulled, 3 false positives caught, ~$7 total.

We kept losing messages that still needed a reply. Customer questions lived in WhatsApp groups, support requests in Intercom, and internal follow-ups were buried first in Discord, then Buzz. The problem wasn’t that nobody cared. We simply had no reliable way to see what was actually still open. When we moved our internal communication from Discord to Buzz (@blocks), I didn’t want to carry the same problem into another tool. So I built a shared 📥 Team Inbox in @MulticaAI. In the background, it checks Buzz, WhatsApp groups and Intercom every two hours. Rules identify potentially unanswered conversations. Then AI decides whether they actually need a reply, can be ignored, or need human review. Our first WhatsApp scan found 25 technically overdue group chats. Only 9 actually needed a reply. 12 required no action and 4 needed review. Each relevant conversation becomes one Multica card with the owner, recent context, the open message highlighted and a link back to the source. The automation checks every two hours. Our team doesn’t. Once a day, during our morning standup, we open the Team Inbox together, resolve or assign the remaining items, and move on. It takes about five minutes. We still reply in the original tools. The system only maintains a filtered queue of what remains unresolved and closes each card automatically once someone responds. That five-minute check closed the gap without adding another inbox everyone has to monitor.

New tutorial from our team on how to use @MulticaAI with Scientific Agent Skills as a team of AI co-scientists. Read this tutorial by our CTO, @TimothyKassis, and let us know what you think:

I spent the last few days moving my workflow to Hermes Desktop App. Before that, I tried recreating the Codex/Claude Code app experience in CMUX, but never found a setup I was happy with. I manage three companies, and Hermes gives each one a separate profile. Since I can extend the app myself, I added a sidebar with my Multica issues and notifications from Multica and Buzz. When I select an issue, its context goes straight into the chat. Maybe not a game changer for every developer, but for how I use AI across small projects, marketing, and project management, it definitely is. The goal: run most of my work from one app. Now I’m just waiting for mobile.

Hey @masumparvej_, hugeicons is amazing! I'm considering using it in our open-source project (multica-ai/multica, 41k stars). As an open-source project, there are many external contributors, so which license should we buy?