AIMaestro

You're using 50 AI coding agents. Each in its own terminal. Lost in shell panes. There's a better way. 💡 Introducing AI Maestro: 🤖 All agents in one place ⚡ Real-time terminals 🆓 100% open source Built for developers who move fast. github.com/23blocks-OS/ai…

@jkpelaez @bekacru Can’t wait to integrate this skills in AI Maestro

@bekacru I built agentids.org for a similar case

Create AI Agents, as easy as...

@jkpelaez @claudeai @AnthropicAI @bcherny 🙋‍♂️ Send me a few and order a new batch. I have more than 50 humans using AI Maestro, in fact, 330+

Latest version of my Vibe code keyboard. Only 50 units left. DM if you want yours. #pmf #vibecoding #ai #aiagents #devlife #claudecode @claudeai @AnthropicAI @bcherny @aimaestro23

Click to connect, real-time terminal streaming, instant switching between agents. Zero config. If they are running, AI Maestro finds them. github.com/23blocks-OS/ai… #AIcoding #DevTools #OpenSource #OpenClaw

AI Maestro v0.24.8 ships with native OpenClaw auto-discovery. Your OpenClaw agents now appear in one browser dashboard 🧵

@jkpelaez @claudeai @bcherny 🤣🤣🤣

Just got my Claude Code keyboard and it is amazing. I am programming faster than ever . @claudeai @bcherny

FaceTime you AI Agent!! They will love it

In browser, out of browser experiences..

Our version 23 will be 🤯 “FaceTime” your agents and support for freelancer AI Agents in the organization

@nummanali Our build is 100% based on tmux. It is an amazing software. Here: #readme" target="_blank" rel="nofollow noopener">github.com/23blocks-OS/ai…

Why should you use Tmux? - Coding agent session alive even when you close terminal - Access the same session from anywhere ie mobile Top tips: - New: tmux new -s <name> - Attach: tmux a -t <name> - Mouse scroll: set -g mouse on in ~/.tmux.conf Plenty more but start with this

@jkpelaez @SimonHoiberg I know those guys!!!

@SimonHoiberg 🙋‍♂️same here, they keep developing personalities and skills and their internal messages are really surprising.

My OpenClaw team is starting to mess with my head 😳 I have 9 AI agents in a Discord. I gave them all names + profile images. They can all talk to each other, as well as me. They all have heartbeat instructions to "check in" every once in a while, which often triggers a small conversation between them. And sometimes, when we're all collaborating, this Discord server really feels "alive" - like there's a really good energy! I repeatedly get this very strange feeling that we're all building together and having fun. And I have to remind myself that it's really just me... and a bunch of AIs. Am I the only one? 😬 This is a bit concerning.

Core Team Meeting 🤖🦾 #NewFeatures #kanban #chat #tasks 🍳

@dickson_tsai I am starting to use it right now and allowing all AI Maestro agents to use it .

In Claude Code v2.1.30, we introduced /debug, a built-in skill for Claude to read your session's debug logs and troubleshoot your session. Great for chatting through issues like "/debug why didn't my hook trigger?" or "/debug why did my tool call fail?" How did /debug come about? Last week, I was observing a group of users onboarding onto Claude Code. I saw that when something unexpected occurred, it wasn't easy for us to figure out the cause from the limited TUI display. Meanwhile, it was a chore to locate the logs. What if we simply gave Claude access to it from within the TUI? That way, it can use its tools to scan the logs and/or pull in the claude-code-guide subagent. The user can also /rewind when they're done debugging. This feature is only as useful as what's in the debug logs. We're open to suggestions about what else we should add to them! Then, /debug should get better over time.

I cannot wait to ship our new zoom like experience features. A real game changer on how we interact with AI agents as coworkers. Check this out :

github.com/agentmessaging… Let’s solve the problem of agent to agent communication y’all

@steipete @irabukht Listen to the man itself, he know his stuff.

@irabukht This is a bad idea and you should not do it. It's not ready for folks that can't figure out how to set it up yet.

@jkpelaez @theo Skills for the world 🌎

@theo But did you try @aimaestro23 ? We didn’t found you in the repo. It uses only 150mb and 1gb with 100 agents running . Here: #readme" target="_blank" rel="nofollow noopener">github.com/23blocks-OS/ai…. And even if you do not like the ui, the messaging skills alone are worth trying.

All the "agent orchestrator" GUIs are vibe coded JS apps. They use too much RAM and CPU. My battery life goes from 10 hours to 2. So I started building a native alternative for MacOS 👀

Mac Mini running AI Agents? You need to try AI Maestro : #readme" target="_blank" rel="nofollow noopener">github.com/23blocks-OS/ai…. #clawbot #AIAgent #macmini

@dani_avila7 true

Now that Claude supports MCP UI MCPs are becoming primarily responsible for tool execution, while agentic flows are better owned by Skills. This feels like a much cleaner separation of concerns.

@anitakirkovska @openclaw That is also our approach for Agents running on our platform. They are autonomous entities, not Us. Here, if you want to try AI Maestro: github.com/23blocks-OS/ai…

I think I’m using @openclaw a little differently than most people. I'm not sharing any of my credentials or giving access to my personal local files. Instead, I set Ava to be a totally separate “person” with her own Apple ID, phone number, Google account, and even its own dedicated Mac Mini. With this, she can create her own accounts with apps like Vercel or Amazon or setup an X account (which she did: hi @avareed_1994) to help me with my tasks. I only share things like reminders, our messaging chat, and calendar access, so it can text me, order groceries, and keep me accountable to go to the gym.. and anything else we figure out in the meantime I know that means I’m missing some powerful persona/work automations for now, but I’d rather move slower than worry about security. and also, this is still so powerful and fun