kingbootoshi

kingbootoshi / ghostbox

Public

Distributed AI agent orchestration - self-evolving ghosts in sandboxed containers

16
1
100% credibility
Found Mar 26, 2026 at 16 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Ghostbox creates persistent AI agents with their own memory, files, and tools that evolve over time, managed via a desktop app, chat, or commands.

How It Works

1
👻 Discover Ghostbox

You find a cool tool for creating personal AI helpers that remember everything they learn.

2
🧙 Run the setup wizard

Follow simple steps to connect your favorite AI services and prepare everything with one command.

3
🚀 Create your first ghost

Give it a fun name and pick a smart brain – your helper comes alive in seconds.

4
💬 Start chatting

Send messages and watch your ghost respond, use tools, and build knowledge as you talk.

5
📱 Open the desktop app

Use the beautiful Mac app to manage ghosts, browse their saved files, and switch between chats.

6
🧠 See it grow

Your ghost saves memories, creates notes, and evolves tools across sessions – nothing is forgotten.

🎉 Persistent AI magic

You now have loyal AI companions that learn from you forever, ready anytime.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 16 to 16 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is ghostbox?

Ghostbox runs persistent AI agents called "ghosts" in isolated Docker containers, each with a git-backed vault for files, warm memory injected into prompts, and deep searchable knowledge. Agents self-evolve by building tools, refining instructions, and saving updates to GitHub, solving the problem of stateless AI chats that forget everything between sessions. Built in TypeScript with Bun for the host orchestrator, it offers CLI commands like `ghostbox spawn researcher` and `ghostbox talk`, plus REST API, Telegram bot, and a native macOS app as a ghostbox app.

Why is it gaining traction?

It stands out with true persistence—vaults survive restarts via git commits—and a two-layer memory system that keeps agents smart without bloating prompts. Developers hook on the self-evolution: ghosts install packages, run servers, and manage their own knowledge via tools like `qmd search`. As a distributed agent orchestration system, it scales multiple agents easily, unlike basic chat wrappers.

Who should use this?

AI engineers building distributed agentic AI workflows, like distributed agent based air traffic flow management sims or research pipelines. Backend devs prototyping distributed agent framework for multi-agent systems, or indie hackers needing a distributed agent runtime with git persistence over distributed GitHub alternatives.

Verdict

Try it for agent experiments—CLI and app make spawning painless—but at 16 stars and 1.0% credibility, it's early alpha with thin docs and no tests. Solid for tinkerers; wait for polish if production-bound.

(187 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.