chaitin/MonkeyCode

企业级 AI 办公工作台

What it is

MonkeyCode (product name: 百智办公 MonkeyAI) is an open-source AI office workbench for teams. You describe a task in plain language, and the AI produces finished work — a presentation, a document, a piece of code, or a design — rather than just giving you advice.

What it does

The README describes three work modes:

  • Office — Make PPTs, write documents, process spreadsheets, organize files. Example prompt: "Turn these materials into a report PPT and summarize the key data from the spreadsheet."
  • Programming — Go from architecture design to writing code and fixing bugs. Example prompt: "Implement this feature from the requirements, add tests, and fix issues found during running."
  • Design — Create posters, graphic designs, interaction designs, and short videos. Example prompt: "Make a poster around the event theme and give matching page designs."

Team management features

The README lists four enterprise capabilities:

  • Shared resources — Teams can share AI models, knowledge bases, skills, experts, MCP connectors, and rules so members reuse what already exists.
  • Usage and billing — Track resource consumption and bill for it.
  • Permissions and auditing — Control who can access which resources, and keep an audit log of usage.
  • Statistics — View team usage data to inform resource planning.

Teams can also plug in their own model services and connect business tools via MCP connectors.

Self-hosting

A one-line install script sets up your own server. Minimum spec: 1 CPU core + 1 GB RAM (adjustable based on team size and workload). Self-hosting means:

  • Your team's data stays on your own server — you control backups, retention, and access.
  • You run it in your own infrastructure and set your own upgrade/maintenance schedule.
  • You can customize the workspace from the open-source code and connect your existing systems.

Clients

Available now: Web, Windows x64, macOS (Apple Silicon and Intel), Linux x64. Coming soon: Android, iOS, Linux ARM64, HarmonyOS desktop.

Pricing

  • Free version — ¥0, up to 10 team seats, no cloud PC, no technical support, basic built-in content.
  • Paid version — Contact sales for pricing; includes cloud PC, technical support, and richer built-in content.

Common questions answered in the README

  • Can non-programmers use it? Yes — describe tasks in natural language, attach files, state the goal and deliverable.
  • Can you bring your own models and tools? Yes — configure model services, connect tools via MCP connectors, and combine with knowledge bases, skills, experts, and rules.
  • Where does data go after self-hosting? Data stays on your server. If you use external models or tools, content is sent to those services — you choose which services to use and set access policies.
  • Can you modify the source? Yes — view, modify, and do secondary development, but distribution must follow the AGPL-3.0 license.

License

GNU Affero General Public License v3.0 (AGPL-3.0).

Demo

A live demo is available: client at https://demo.work.baizhi.cloud/ (log in with email verification code) and an admin demo at https://admin.demo.work.baizhi.cloud/ with credentials admin@baizhi.cloud / work.baizhi.cloud.

Note

The repository also retains the original MonkeyCode project introduction in separate files (README.monkeycode.md and readme.cn.md), but the main README describes the current MonkeyAI product.

Related

  • Project
    whitelonng/mancode

    mancode is a workflow orchestration tool that adds a governance layer to AI coding agents, ensuring they follow project standards, reuse existing code, and maintain privacy while working.

  • Project
    aicodeking/pandacode

    PandaCode is an open‑source native desktop app that lets you run many AI coding agents (Claude Code, Codex, Kimi Code, etc.) alongside Git, a terminal and a browser. It offers a Kanban task board, agent‑team orchestration, bot integrations, local/cloud sandboxes, one‑click deploys and deep Git support, all stored locally with no telemetry.

  • Project
    PackmindHub/packmind

    A centralized engineering playbook that synchronizes coding standards and instructions across multiple AI coding agents and repositories.

  • Project
    landingbj/LinkMind

    An enterprise-grade multimodal AI middleware that provides a unified API layer for managing multiple model providers, RAG, and agent runtimes with centralized routing and governance.

  • Project
    zhikunqingtao/zhikuncode

    An open-source, self-hosted AI programming assistant that enables full development workflow control via a web browser, featuring advanced context management and runtime verification.