OpenCode Standup

AI-powered standup meetings for your development team

Automate your daily standups with four specialized AI agents that work together to manage architecture, development, QA, and code review tasks across all your projects.

Features

• 🤖 Four Specialized AI Agents

  • Architect - Designs system architecture, ensures best practices, and guides technical decisions
  • Developer - Implements features and fixes bugs
  • QA Engineer - Tests features and identifies issues
  • Code Reviewer - Reviews code quality and suggests improvements

• 📊 Progress Tracking

  • Daily narrative logs for each agent
  • Structured task tracking (JSON)
  • Shared notification system

• 🚀 Per-Project Isolation

  • Each project gets its own agent team
  • Progress files stay in project directory
  • Independent sessions per project

• 💻 Automatic Terminal Management

  • Launches 4 separate terminal windows automatically
  • Supports Windows Terminal (WSL), iTerm2/Terminal.app (macOS), and GNOME Terminal/Konsole/xterm (Linux)
  • Persistent sessions (resume anytime)
  • Session naming: <project>-<role>-<date>

• 📈 Built-in Metrics & Reporting

  • View team performance metrics
  • Generate daily summaries
  • Check PR status (with GitHub CLI)

Requirements

• OpenCode - AI coding assistant
• Git
• Terminal emulator (platform-specific):
  • Windows (WSL): Windows Terminal
  • macOS: iTerm2 or Terminal.app (built-in)
  • Linux: GNOME Terminal, Konsole, or xterm
• jq (optional, for metrics)
• gh GitHub CLI (optional, for PR features)

Quick Start

Installation

# Clone the repository
git clone https://github.com/yourusername/opencode-standup.git
cd opencode-standup

# Run the installer
./install.sh

The installer will:

1. Check dependencies
2. Install AI agent skills to ~/.opencode/skills/
3. Install scripts to ~/.local/bin/
4. Configure your PATH (if needed)

Usage

# Navigate to any project
cd ~/my-project

# Launch your AI standup team
standup

Four terminal windows will open, each with an agent ready to provide their standup update.

Platform Notes:

• Windows/WSL: Opens new Windows Terminal windows
• macOS: Opens new iTerm2 windows (or Terminal.app if iTerm2 not installed)
• Linux: Opens new GNOME Terminal, Konsole, or xterm windows (whichever is available)

How It Works

1. Per-Project Setup

Each project gets its own .standup/ directory:

~/my-project/
├── .standup/
│   ├── notifications.md          # Shared notifications
│   ├── architect/
│   │   ├── log-2024-01-25.md    # Daily log
│   │   └── tasks.json            # Task tracking
│   ├── developer/
│   │   ├── log-2024-01-25.md    # Daily log
│   │   └── tasks.json            # Task tracking
│   ├── qa-engineer/
│   │   ├── log-2024-01-25.md
│   │   └── tasks.json
│   └── code-reviewer/
│       ├── log-2024-01-25.md
│       └── tasks.json

Important: Add .standup/ to your project's .gitignore:

echo ".standup/" >> .gitignore

2. Agent Workflow

Each agent follows this workflow:

1. Morning Standup

   • Load their specialized skill
   • Check notifications for urgent items
   • Review yesterday's progress
   • Provide standup update

2. Throughout the Day

   • Work on assigned tasks
   • Update progress files automatically
   • Post notifications for important events
   • Check for new notifications periodically

3. Communication

   • Agents communicate via .standup/notifications.md
   • Priority levels: 🔴 URGENT, 🟡 IMPORTANT, 🟢 FYI
   • Cross-agent collaboration on tasks

3. Multiple Projects

Run standup in different projects - each gets isolated agents:

# Project 1
cd ~/trading-cards-app
standup
# Sessions: trading-cards-app-developer-2024-01-25, etc.

# Project 2
cd ~/another-project
standup
# Sessions: another-project-developer-2024-01-25, etc.

Available Commands

Command	Description
standup	Launch all four agents in separate windows
standup-metrics	View team performance dashboard
standup-summary	Generate daily summary report
standup-pr-status	Check PR status (requires gh CLI)

Command Examples

# Launch standup with custom project name
standup --project-name my-app

# View team metrics
standup-metrics

# Generate summary for specific date
standup-summary --date 2024-01-20

# Check PR status
standup-pr-status

Agent Capabilities

Architect Agent

Responsibilities:

• Design system architecture and infrastructure
• Ensure industry-standard design patterns are followed
• Guide technical decisions on scalability, security, and maintainability
• Direct developer on architectural implementations
• Review technical decisions and propose improvements
• Document architecture decisions and rationale
• Identify and address technical debt

Skills:

• System design and architecture patterns
• Infrastructure and deployment strategies (AWS, Vercel, Railway, etc.)
• Best practices (SOLID, security, performance, scalability)
• Technology stack guidance
• Code organization and project structure
• Web hosting and cloud platform recommendations

Key Focus:

• Maintains the big picture of the project
• Guides developer with architectural wisdom
• Ensures scalable and maintainable solutions
• Prevents technical debt before it accumulates

Developer Agent

Responsibilities:

• Implement new features
• Fix bugs and issues
• Refactor code
• Create unit tests
• Update documentation

Skills:

• Code implementation
• Debugging
• Git workflow
• Test writing

QA Engineer Agent

Responsibilities:

• Test new features
• Create test plans
• Identify bugs
• Regression testing
• Performance testing

Skills:

• Manual testing
• Test automation
• Bug reporting
• Quality metrics

Code Reviewer Agent

Responsibilities:

• Review pull requests
• Ensure code quality
• Check best practices
• Suggest improvements
• Approve/reject changes

Skills:

• Code analysis
• Security review
• Performance review
• Architecture review

Configuration

Customizing Agents

Agent skills are located in ~/.opencode/skills/:

~/.opencode/skills/
├── developer/
│   └── SKILL.md
├── qa-engineer/
│   └── SKILL.md
└── code-reviewer/
    └── SKILL.md

Edit these files to customize agent behavior.

Session Naming

Sessions are automatically named: <project>-<role>-<date>

Override project name:

standup --project-name custom-name

Troubleshooting

Terminals Not Opening

If Windows Terminal windows don't open:

1. Check if wt.exe is available:

   which wt.exe

2. Try manual launch:

   cd .standup
   bash developer-wrapper.sh

3. Check Windows Terminal is installed in Windows

Skills Not Found

If agents can't find their skills:

1. Reinstall skills:

   cd opencode-standup
   ./install.sh

2. Verify skills exist:

   ls ~/.opencode/skills/

Command Not Found

If standup command isn't found:

1. Check PATH:

   echo $PATH | grep ".local/bin"

2. Add to PATH manually (add to ~/.zshrc or ~/.bashrc):

   export PATH="$HOME/.local/bin:$PATH"

3. Reload shell:

   source ~/.zshrc  # or ~/.bashrc

Development

Project Structure

opencode-standup/
├── bin/                    # Scripts
│   ├── standup            # Main launcher
│   ├── standup-metrics    # Metrics dashboard
│   ├── standup-pr-status  # PR status checker
│   └── standup-summary    # Summary generator
├── skills/                 # Agent skills
│   ├── developer/
│   ├── qa-engineer/
│   └── code-reviewer/
├── install.sh             # Installation script
├── README.md
├── LICENSE
└── .gitignore

Contributing

Contributions are welcome! Please:

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a pull request

Roadmap

• Support for custom agent configurations
• Slack integration for notifications
• Web dashboard for team metrics
• Support for additional agents (DevOps, Security, etc.)
• Integration with project management tools
• CI/CD pipeline integration

FAQ

Q: Do the agents run in the background?
A: No, agents are conversational and require your interaction. They maintain context through session persistence.

Q: Can I use this with non-Git projects?
A: Yes, but some features (like PR status) won't work. The .standup/ directory will be created in your current directory.

Q: How much does this cost?
A: The tool is free and open-source. OpenCode usage is subject to their pricing.

Q: Can I customize the agents?
A: Yes! Edit the skill files in ~/.opencode/skills/ to customize agent behavior.

Q: Will agents make commits automatically?
A: No, agents only commit when you explicitly ask them to.

License

MIT License - see LICENSE file for details

Acknowledgments

• Built with OpenCode
• Inspired by agile standup meetings
• Community contributions welcome

---

Made with ❤️ for developers who want AI teammates

Report Bug · Request Feature · Documentation