English | 中文

open-teams

open-teams — AI collaboration workspace template for Cursor, Copilot & Claude Code.
Git-native AGENTS.md, MEMORY.md, Skills & task plans turn AI from a one-shot chat into a first-class teammate.

Quick Start • Why? • Features • Structure • Comparison • Roadmap • Discussions

Why open-teams?

Your team loves AI coding tools. Cursor, GitHub Copilot, Claude Code — everyone's using them. Productivity is up. But after a few sprints, you notice something:

• Context vanishes. Every new AI chat starts from zero. Yesterday's architecture decision? Gone.
• Standards diverge. Five team members, five different Cursor rules. Code review is a style-war zone.
• Knowledge leaks. Wang's killer prompt lives in his chat history. Li's debugging trick stays in her DMs. Nobody else benefits.
• Onboarding is manual. Every new hire re-teaches AI from scratch. Two weeks before they're productive.

The problem isn't the AI. It's the layer between the AI and the team.

Who it's for: engineering teams that want a shared AI workspace template — one Git repo for team constitution (AGENTS.md), long-term memory (MEMORY.md), modular Skills, and task plans — without locking into a single AI vendor.

open-teams is that layer — a lightweight, Git-native workspace template that gives your AI:

• 🧠 Memory — so it remembers what happened last session (MEMORY.md + daily notes)
• 📋 Shared Standards — so everyone's AI follows the same AGENTS.md rules
• 🔌 Modular Skills — reusable workflows for code review, API design, architecture review
• 📝 Knowledge Capture — so insights become team assets, not lost chat history

"Not a new AI tool. An operating system for how your team collaborates with AI."

Architecture Overview

┌──────────────────────────────────────────────────────────┐
│                  Your AI Coding Tool                      │
│         (Cursor / Copilot / Claude Code / ...)            │
└─────────────────────┬────────────────────────────────────┘
                      │ reads & writes
                      ▼
┌──────────────────────────────────────────────────────────┐
│                   open-teams Workspace                     │
│                                                           │
│  ┌─────────────┐  ┌──────────────┐  ┌─────────────────┐  │
│  │  AGENTS.md   │  │  MEMORY.md   │  │  TOOLS.md       │  │
│  │  Team        │  │  Long-term   │  │  Environment-   │  │
│  │  Constitution│  │  Memory      │  │  specific Notes │  │
│  └─────────────┘  └──────────────┘  └─────────────────┘  │
│                                                           │
│  ┌───────────────────────────────────────────────────┐   │
│  │  Skills (Modular AI Workflows)                     │   │
│  │  ┌──────────┐ ┌──────────┐ ┌──────────────────┐   │   │
│  │  │code-     │ │api-design│ │architecture-     │   │   │
│  │  │review/   │ │-review/  │ │review/           │   │   │
│  │  │SKILL.md  │ │SKILL.md  │ │SKILL.md          │   │   │
│  │  │rules/    │ │rules/    │ │rules/            │   │   │
│  │  │examples/ │ │examples/ │ │examples/         │   │   │
│  │  └──────────┘ └──────────┘ └──────────────────┘   │   │
│  └───────────────────────────────────────────────────┘   │
│                                                           │
│  ┌─────────────┐  ┌──────────────┐  ┌─────────────────┐  │
│  │  docs/       │  │  memory/      │  │  task-plans/   │   │
│  │  Team Docs   │  │  Daily Notes  │  │  Sprint Plans  │   │
│  └─────────────┘  └──────────────┘  └─────────────────┘  │
└──────────────────────────────────────────────────────────┘
                      │ version controlled
                      ▼
┌──────────────────────────────────────────────────────────┐
│                       Git Repository                       │
│    AI rules, memory, and skills — reviewed in PRs like code │
└──────────────────────────────────────────────────────────┘

Key insight: AI reads Markdown natively. Git tracks everything. Your AI's "brain" gets the same rigor as your codebase.

Features

Quick Start

Prerequisites

• Git installed
• Any AI coding tool (Cursor, Copilot, Claude Code, etc.)
• 5 minutes

Setup

This repository is a monorepo template with two complete workspaces:

# 1. Clone the template (keep as read-only reference)
git clone https://github.com/struggling-bird/open-teams.git open-teams-template

# 2. Open the English workspace (or use zh/ for Chinese)
cursor open-teams-template/en

# 3. Adopt into your business workspace (run from the repo root)
python3 open-teams-template/scripts/adopt_workspace.py \
  --template-root open-teams-template \
  --target /path/to/your-workspace \
  --project-slug my-app \
  --project-name "My App" \
  --source-path /path/to/source \
  --default-branch main \
  --adoption-mode ai-guided \
  --workspace-language en

# 4. Open the target folder in your AI tool and finish MEMORY.md / AGENTS.md

Use --workspace-language zh to copy the Chinese workspace instead. Full walkthrough: en/QUICKSTART.md · zh/QUICKSTART.md · en/docs/getting-started.md.

That's it. Your AI now:

• ✅ Knows your project structure and tech stack (from AGENTS.md)
• ✅ Remembers past decisions (from MEMORY.md)
• ✅ Follows shared review standards (from skills/)
• ✅ Writes daily collaboration notes (in memory/)

Try your first Skill

# Open your AI tool in the adopted workspace (or en/), then prompt:
"Review the latest commit using the code-review skill"

The AI reads skills/code-review/SKILL.md, loads the checklist and rules, and delivers a structured review — every team member gets the same quality standard.

English retention path

1. This README + en/QUICKSTART.md
2. en/docs/getting-started.md
3. Governance contracts: en/AGENTS.md, en/workspace-assets-index.md, en/task-completion-checklist.md, en/MEMORY.md
4. Chinese docs: zh/README.md

Project Structure

open-teams/
├── README.md                  # You are here — GitHub landing (English)
├── LICENSE · CHANGELOG.md · CONTRIBUTING.md · SECURITY.md
├── .github/                   # CI, issue & PR templates
├── scripts/adopt_workspace.py # Copy en/ or zh/ into a new workspace
├── assets/logo.svg
│
├── en/                        # Complete English workspace ← open this in Cursor
│   ├── AGENTS.md              # Team constitution — AI reads this first
│   ├── MEMORY.md              # Long-term team memory
│   ├── QUICKSTART.md          # Dialogue-guided adoption
│   ├── skills/                # Modular AI capabilities
│   │   ├── _workflow/         # Cross-project process gates
│   │   ├── code-review/
│   │   ├── api-design-review/
│   │   ├── architecture-review/
│   │   └── onboarding/
│   ├── docs/                  # Architecture, specs, getting-started
│   ├── workspace-config/
│   ├── task-plans/ · change-history/ · memory/ · sources/
│   └── scripts/               # validate / health-check / cleanup
│
└── zh/                        # Complete Chinese workspace ← 中文请打开这里
    ├── AGENTS.md
    ├── MEMORY.md
    ├── QUICKSTART.md
    ├── skills/ · docs/ · workspace-config/ · …
    └── …

Comparison

open-teams vs. Raw AI Chat vs. Traditional Project Management

open-teams vs. Competitors (Detailed)

Roadmap

💡 Have a feature idea? Open a Discussion.

Contributing

open-teams is community-first. Whether you're fixing a typo or proposing a new Skill, you're welcome.

Ways to Contribute

• ⭐ Star the repo — it helps more people find us
• 🐛 Report bugs — Open an Issue
• 💡 Suggest features — Start a Discussion
• 🔧 Submit PRs — See CONTRIBUTING.md
• 📖 Improve docs — Keep en/ and zh/ in sync when changing governance or install guides
• 🧩 Share Skills — Built a useful Skill? Share it with the community
• 🌍 Translate — Help localize docs for global teams

Good First Issues

Check out issues labeled good first issue — we'll help you through your first PR.

Community

• 💬 GitHub Discussions — Questions, ideas, RFCs
• 📢 Follow for updates (links coming soon)

Star History

License

MIT © struggling-bird

_{Built with ❤️ by developers who believe AI should augment teams, not replace them.
⭐ Star us on GitHub}