RunMaestro.ai

Maestro is a cross-platform desktop app for orchestrating your fleet of AI agents. Set them loose on complex tasks, check in from your phone, and let them work while you sleep. Supports Claude Code, Codex, and OpenCode. Leveraging all your existing skills, MCP, and subagents.

@ArtemXTech ...or... use a delightful open source UX: github.com/RunMaestro/Mae…

@PaulRBerg Solve this and a myriad of other UX desires with a beautiful cross-platform desktop wrapper. Open source, free (as in beer): github.com/RunMaestro/Mae…

This is the most annoying thing in Claude Code. Hiding raw text when you paste more than 4 lines. Terrible UX decision.

That is the pinnacle of the UX today, but there are so many other features that will make managing your dozens of agents far easier. Group chat for example allows you to have agents, backed by different providers (Codex, Claude, etc.) work with one another. We're actively working on Maestro Cue at the moment, that brings triggers into the system. Think heartbeats, cron jobs, file/directory triggers, Github issue/PR triggers, etc.

@emelle0x @RunMaestroAI It's always fun to see creative takes on beloved characters! Ralph Wiggum brings such a nostalgic charm; I can only imagine how playful the loop GUI must be!

So @RunMaestroAI is... a Ralph Wiggum loop GUI? runmaestro.ai

@garrytan That's awesome. Now to upgrade your interfacing experience! github.com/RunMaestro/Mae… All your skills, subagents, MCPs and other customizations will convey. Free. Open source. Cross platform. Actively developed. Instantly addictive.

I've been having such an amazing time with Claude Code I wanted you to be able to have my *exact* skill setup: Introducing gstack, which you can install just by pasting a short piece of text into your Claude code

Apparently, at our peak, we were the #2 on GitHub! trendshift.io/repositories/2… Next goal... hitting #1

Hacking and vibe coding chat with @pedramamini and @gregcmartin on the @ghostsecurityhq podcast: youtube.com/watch?v=kBrHEa…

Adding Maestro to my AI development workflow and agents. @RunMaestroAI #AI #Agents #AgenticAI so far, so good 😊

Also coming in 0.16.0-RC is the first release of Maestro Cue, a pipelining system allowing you to assign your agents to a variety of triggers including: time based, file system monitoring, Github, tasks, other agents and more.

0.16.0-RC will be available soon, featuring an upgraded terminal (xterm.js), in-tab, side-by-side with your AI chat and file preview tabs. Spawn multiple terminals, rename them, content persists between focus shift.

@adocomplete We've had this feature since v0.2.0 (November 28th 2025) docs.runmaestro.ai/remote-access

Run /remote-control to continue building on the go. Remote control in Claude Code is probably my favorite feature that shipped this year.

Amazing idea @pedramamini 👌💪 "Maestro Symphony is a new way to contribute to open source projects. Pick an issue from the Symphony board that interests you, run it locally, and watch your contribution come to life."

@fullpatstack Hah! Those early achievements come easy... gets real hard real fast. What are you hacking on?

@RunMaestroAI is fun. but i don't really know what I am even doing lol

Finally out! So many changes. Much of them internal refactoring. We'll be iterating more quickly again. Read the full release notes: github.com/RunMaestro/Mae… Feedback welcome!

@om_patel5 That's what our desktops looked like before we wrote Maestro. If you're working at that level of parallelism, give this open source desktop a try and please provide feedback.

this guy just hired 25 founding engineers (it's 25 sessions of claude code)

@coreyganim If you like that, you should try Maestro. Similar concepts, way more advanced. As addictive as it is powerful. Open source.

Best Cowork setup guide I've seen. How to fully implement in under 15 minutes: Setup (do once) □ Download Claude Desktop → claude. com/download □ Sign in → Click "Cowork" tab at top □ Create folder: ~/Claude-Workspace/ □ Inside it, create: /context, /projects, /outputs Context files (this is the leverage) □ Create context/about-me.md → Your role, what you do, what success looks like □ Create context/brand-voice.md → Paste 2-3 examples of YOUR writing → List phrases you hate ("leverage", "I hope this finds you well") □ Create context/working-preferences.md → "Ask questions before starting" → "Show plan before executing" → "Never delete without confirmation" Plugins □ Click Customize → Browse plugins □ Install: Productivity (everyone needs this) □ Install: One for your role (Sales, Marketing, Finance, etc.) □ Type "/" to see available commands Connectors □ Settings → Connectors □ Connect your top tool: Slack, Google Drive, or Notion First task □ Paste this to verify setup works: "Read all files in context/. Tell me what you know about me and how I prefer to work." Then try: □ "Organize this folder by type and date" □ "Create a report from these scattered notes" □ "Draft a doc based on this template" The compounding trick: After every session where output missed the mark, update your context files. Hope this helps!

@oliveiraacc @cristofrcharles @nyk_builderz There's so many of these tools. To each their own!

@cristofrcharles @nyk_builderz Looks like @RunMaestroAI just not as good.

We just open-sourced Mission Control — our dashboard for AI agent orchestration. 26 panels. Real-time WebSocket + SSE. SQLite — no external services needed. Kanban board, cost tracking, role-based access, quality gates, and multi-gateway support. One pnpm start, and you're running. github.com/builderz-labs/…

@softcoda @FactoryAI It's as easy to use as it's addictive.

@RunMaestroAI @FactoryAI Wowww….coool. This is an amazing feature and an effective pathway to full autonomous project development (from just code development).

Droids can now pursue goals autonomously over multi-day horizons. You describe what you want, approve the plan, and come back to finished work. We call these Missions.

Latest 0.15.0 RC allows you to spin out Auto Run docs in a worktree. Next up... automatic parallelization of docs across worktrees!

The Mozilla 0DIN.ai team just published my guest blog post on the art of unattended AI coding sessions. I cover some of the history before taking a dive into Maestro, an open source app with an active community of over 200 Discord users and over 2,300 stars as of this time of writing. In the piece, I explore how to move past "babysitting" AI. Showing how persistent iteration beats one-shot perfection, and how we’ve managed to push unattended runtimes from hours to days. If you’re interested in understanding the pillars of autonomous automation, check out the full post here: 0din.ai/blog/while-tru…

@EnoReyes @FactoryAI You've got some serious fans, because you'll notice that Factory came before Gemini and Qwen3!

@RunMaestroAI @FactoryAI This is awesome!