Tutti: A Multi‑Agent Workspace That Seamlessly Connects Claude, Codex and More

Tutti is an open‑source, OS‑like workspace that lets multiple AI agents such as Claude Code, Codex and other tools share context, files and tasks in real time, eliminating the need to repeatedly re‑explain project background, as demonstrated by building a 2026 World Cup tracker app.

IT Services Circle
IT Services Circle
IT Services Circle
Tutti: A Multi‑Agent Workspace That Seamlessly Connects Claude, Codex and More

Tool Overview

Tutti is an open‑source, multi‑agent real‑time shared workspace that bridges context, applications, tasks, and files. The UI mimics an operating system with a dock of quick‑launch apps. It can integrate Claude Code, Codex, and placeholders for future agents such as Hermes, Gemini, and OpenClaw. Built‑in apps include product prototype design, AI document generation, PPT creation, and image generation.

Context Transfer Mechanism

Within Tutti, typing @ followed by an agent name in the input box injects the entire shared workspace into that agent. The injected payload contains the full conversation history, requirement text, agent replies, file diffs, and all code changes. Consequently, a new agent (e.g., Codex) can continue work without the user manually summarising or re‑entering the project state.

Hands‑On Demonstration: 2026 World Cup Tracker

Requirement gathering – Claude Code is launched inside Tutti and given a prompt (shown in the original screenshots) to produce a clear product‑requirements document (PRD). The PRD is stored in the shared space.

Prototype generation – The built‑in “Product Prototype Design” app is invoked by typing @产品原型设计 in the Claude Code conversation. A single‑sentence request generates a UI prototype, which can be previewed instantly in Tutti’s embedded browser and edited on the fly.

Code development – Claude Code reads the design file directly from the shared workspace and starts implementing the pages. Because the design artefact already resides in the same environment, no manual download or screenshot step is required.

Image generation – When a page lacks a venue illustration (Claude Code cannot generate images), the user @‑mentions the “AI Canvas” app. The app runs GPT‑Image 2, creates the required image, and inserts it into the code base without leaving Tutti.

Agent hand‑off to Codex – Codex is @‑mentioned, receives the full context (requirements, prototype, code, and generated image), reviews the current code, and applies the new image. No additional background explanation is needed.

Key Observations

Agents share not only textual context but also artefacts produced by other apps (design files, images, diffs).

The @‑mention syntax acts as a lightweight orchestration layer, allowing a single command to switch execution from one agent to another while preserving project state.

Typical multi‑agent workflows (requirement → design → coding → asset creation) that normally require repeated copy‑pasting or manual hand‑offs are collapsed into a continuous flow within the same workspace.

The approach works with existing Claude or Codex subscriptions; no extra licensing cost is introduced.

Repository

GitHub project address: https://github.com/tutti-os/tutti

Original Source

Signed-in readers can open the original source through BestHub's protected redirect.

Sign in to view source
Republication Notice

This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactadmin@besthub.devand we will review it promptly.

open sourceMulti-AgentClaudeCodexagent collaborationAI workspaceTutti
IT Services Circle
Written by

IT Services Circle

Delivering cutting-edge internet insights and practical learning resources. We're a passionate and principled IT media platform.

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.