sonichi/sutando

My AI Stand. Realtime by day, rewriting itself by night. Summon my AI superpower.

What it solves

Sutando is a personal AI agent designed to act as a "Stand" (a personal spirit) that manages your digital life across macOS. It solves the problem of fragmented AI interactions by providing a single, persistent agent that can see your screen, join meetings, make phone calls, and autonomously improve its own code and capabilities while you are away.

How it works

Sutando operates through a combination of four primary processes:

  • Core Agent: Powered by Claude Code or Codex CLI, this agent has full system access and executes tasks via a scheduled proactive loop (every 15-30 minutes) and a persistent task watcher.
  • Voice Agent: Uses Gemini Live via WebSockets to provide real-time voice interaction in the browser.
  • Conversation Server: Handles inbound and outbound phone calls via Twilio WebSockets.
  • Web Client: Provides a browser-based UI for interacting with the agent.

For instant actions (like describing a screen), the voice and phone agents use "inline tools." For more complex work, they write tasks to a tasks/ directory, which the Core Agent then processes and writes results to a results/ directory.

Who it’s for

It is built for macOS users who want a highly integrated, autonomous AI assistant that can handle system-level tasks, manage communications (Zoom, Google Meet, Phone, Telegram, Discord), and evolve its own functionality through an autonomous build loop.

Highlights

  • Multimodal Interaction: Supports voice, vision (screen capture), and text across multiple channels including phone, Telegram, and Discord.
  • Autonomous Evolution: Runs a build loop to monitor its own health, detect user patterns, and ship its own code to add missing capabilities.
  • Meeting & Call Management: Can join Zoom/Google Meet calls, take screenshots to identify participants, and make concurrent phone calls.
  • Cross-Machine Scaling: Can coordinate with other Mac machines to migrate services and share a consistent identity and memory.
  • Local First: Runs entirely on the user's Mac with no remote control plane.

Related

  • Project
    AgentiLoop/Agent

    A native macOS app that provides an autonomous AI agent capable of controlling the desktop, managing codebases, and automating system tasks across 18 LLM providers.

  • Project
    Sentient-OS-Labs/sentient-os

    An on-device AI operating system for macOS that builds a private, local knowledge base of your digital life to power proactive agents and computer-use automation.

  • Project
    jamesrochabrun/AgentHub

    A macOS desktop hub for managing multiple AI coding agent sessions (Claude Code and Codex) with integrated terminal, GitHub support, and live previews.

  • Project
    Kocoro-lab/Kocoro

    Kocoro is an open‑source macOS AI‑agent daemon (`shan`) that lets language‑model agents read/write files, control apps, run shell commands, automate browsers, and integrate with Slack/Telegram. It provides a CLI/TUI, HTTP API, MCP streaming, and a real‑time voice front‑brain. Install via npm, a script, or Go build; configure a Shannon Cloud or self‑hosted gateway for LLM completions. The daemon handles permissions, per‑agent memory, scheduling, and extensible tools, while a separate closed‑source GUI (Kocoro Desktop) offers a polished interface. Licensed under MIT.

  • Project
    pocketpaw/pocketpaw

    A self-hosted AI agent that runs locally on your machine, providing a native desktop app and integration with multiple messaging platforms like Discord and Slack.