dream-num/univer-workspace

An open-source Office workspace where people and AI agents create, collaborate, and review together.

What it solves

Univer Workspace provides a collaborative office environment where humans and AI agents can co-create and review documents. It solves the problem of isolating AI-generated changes from stable content, allowing agents to propose edits in a sandbox before a human approves them, and enables the creation of spreadsheet-based mini-apps where web-based controls are bound directly to spreadsheet cells.

How it works

The system consists of three primary entry points connected to a central Workspace Server:

  • Workspace Browser: The main interface for humans to edit documents and review agent changes.
  • Workspace Agent: A local application for chatting with an AI agent to discuss documents and manage reviews.
  • Workspace CLI: A tool for AI agents to programmatically load, edit, and validate content.

To ensure stability, it uses a "Worktree" workflow: agents make changes in an isolated draft, mark it as ready, and a human then reviews and merges those changes into the main document (the trunk).

Who it’s for

  • People: Users who need to organize content in shared spaces and co-edit office documents (Sheets, Docs, Slides, etc.).
  • AI Agents: Developers building agents that need to create rich office content via API, render PDFs/screenshots, or run layout checks.
  • Operators: System administrators who need to deploy a self-hosted collaboration workspace with controlled data and authentication.

Highlights

  • Worktree Review Workflow: Prevents AI agents from directly overwriting live data by requiring human approval for merges.
  • Multi-Artifact Support: Integrated environment for Sheets, Docs, Slides, relational tables (Bases), and canvases (Boards).
  • Spreadsheet Mini-Apps: Ability to generate interactive dashboards where UI elements are bound to spreadsheet cells.
  • Agent-Ready Tooling: Includes a dedicated CLI and Facade API for structured AI interaction with office documents.

Related

  • Project
    microsoft/kernel-memory

    Kernel Memory is a multi-modal AI service for efficient dataset indexing and Retrieval Augmented Generation (RAG), providing tools for data ingestion pipelines and natural language querying with citations.

  • Project
    daodao97/chatmcp

    A cross-platform AI chat client that integrates with the Model Context Protocol (MCP) to connect LLMs with external data sources and tools.

  • Project
    tsingyuai/scientify

    An end-to-end autonomous AI research system that autonomously conducts literature reviews, generates hypotheses, and performs experiments to produce SOTA scientific results.

  • Project
    777genius/agent-teams-ai

    A desktop orchestration layer for AI agent teams that allows multiple agents from different providers to collaborate autonomously on tasks via a visual Kanban board.