Maestri
Maestri is a native macOS application that provides an infinite canvas for orchestrating AI coding agents. It transforms the traditional terminal into a visual workspace where developers can position terminal nodes, notes, and sketches freely. The app allows users to connect multiple agents like Claude Code, Codex, and Gemini, enabling them to collaborate across harnesses through PTY orchestration. It features 'Ombro,' an on-device AI companion powered by Apple Intelligence that monitors and summarizes agent activity. Built with SwiftUI and Metal, Maestri emphasizes high performance and privacy, operating with zero cloud dependency and no telemetry.
Infinite visual canvas for orchestrating and connecting multiple AI coding agents on macOS.
- Managing multiple AI agents simultaneously for complex coding projects
- Visualizing system architecture alongside live terminal agents
- Using on-device AI to summarize progress while away from the keyboard
- Switching between different project contexts using dedicated canvases
- Download and launch the app on a Mac with Apple Silicon. Drag and drop terminal nodes onto the infinite canvas to start agents like Claude Code or Gemini. Connect terminals with lines to enable agent collaboration. Use the built-in sketching tools and sticky notes to document your workflow. Access Ombro for automated task summaries.
The AI database client for Mac
API integration testing that remembers what breaks
Caps Lock keeps your Mac awake, even with the lid closed
Sync your mailbox with your issue tracker
See it once. Ask forever.
The open source Ahrefs alternative
Your time travel debugger
Control panel for AI agents on your lock screen
Catch API drift before your customers do
A free, open source, Descript alternative. Runs in-browser.
Full-Stack Platform for Training Small Language Models
Instant, hardware Isolated Sandboxes for AI agents
The web performance & security company
A modern, visual file manager for the terminal
AI platform for deep video understanding
Control every aspect of model training and fine-tuning
The easiest way to deploy a managed agent.
Fastest LLM runtime on Apple Silicon
