AITeamForge - AI-powered multi-team development infrastructure
This Homebrew tap provides the
aiteamforgeformula for installing and managing the AITeamForge environment on macOS - a comprehensive AI-assisted development platform with specialized teams, visual kanban management, and multi-machine coordination.
AITeamForge is a comprehensive development environment designed for AI-assisted development with multiple specialized teams.
Separate teams (iOS, Android, Firebase, etc.) with distinct AI agents, kanban boards, and workflows.
aiteamforge start ios # Start iOS team
ios-picard # Launch Captain Picard agent
kb-list # View iOS kanban boardStar Trek-inspired visual kanban board with real-time updates and automatic agent tracking.
kb-add "New feature" # Add kanban item
kb-move XIOS-0042 in_progress # Move to in progress
kb-run XIOS-0042 # Create worktree and start workAccess in browser: http://localhost:8082
AI pair programming with team-specific personas that auto-track work in kanban.
ios-picard # Lead Feature Developer
ios-beverly # Bugfix Specialist
ios-data # Testing & QAEach agent has a unique personality and specialization based on Star Trek characters.
Advanced git workflow automation for parallel development.
wt-create feature/new-ui # Create worktree
wt-list # List worktrees
wt-remove feature/new-ui # Clean up after mergeMulti-machine coordination for distributed development across multiple Macs.
# View all machines, agents, and kanban state
open http://localhost:3000iTerm2 integration with automated window/tab management (optional).
aiteamforge start ios
# Automatically opens terminals, configures environmentsGet up and running in under 5 minutes:
# 1. Add tap and install
brew tap DoubleNode/aiteamforge
brew install aiteamforge
# 2. Run interactive setup wizard
aiteamforge setup
# 3. Verify installation
aiteamforge doctor
# 4. Start your environment
aiteamforge start iosThat's it! Your AI-powered development environment is ready.
- macOS Big Sur (11.0) or later
- Homebrew - Install here if you don't have it
Step 1: Add the AITeamForge tap
brew tap DoubleNode/aiteamforgeStep 2: Install AITeamForge
brew install aiteamforgeThis installs the framework and required dependencies (Python 3, Node.js, jq, GitHub CLI, Git).
Step 3: Run the Setup Wizard
aiteamforge setupThe interactive wizard guides you through:
- ✓ Checking dependencies
- ✓ Installing missing tools (if needed)
- ✓ Selecting teams (iOS, Android, Firebase, etc.)
- ✓ Configuring features (LCARS Kanban, Fleet Monitor, etc.)
- ✓ Setting up shell environment
- ✓ Installing system services
Typical setup time: 5-10 minutes (depending on how many dependencies need installation).
Step 4: Restart Your Terminal
source ~/.zshrcStep 5: Verify Everything Works
aiteamforge doctoraiteamforge setup # Run interactive setup wizard
aiteamforge doctor # Health check and diagnostics
aiteamforge status # Show current environment status
aiteamforge start # Start aiteamforge environment
aiteamforge stop # Stop aiteamforge environment
aiteamforge upgrade # Upgrade components
aiteamforge help # Show help informationFor iOS Development:
# Install and select iOS team
aiteamforge setup # Choose "iOS Development"
# Start iOS environment
aiteamforge start ios
# Use iOS agents
ios-picard # Captain Picard - Lead Feature Developer
ios-beverly # Dr. Crusher - Bugfix Specialist
# Manage tasks
kb-list # List kanban items
kb-add "New feature"For Multi-Platform Development:
# Install iOS, Android, and Firebase teams
aiteamforge setup # Choose multiple teams
# Start all teams
aiteamforge start
# Each team has its own kanban board and agentsFor Multi-Machine Setup:
# On main machine (server)
aiteamforge setup # Enable "Fleet Monitor" as server
# On secondary machines (clients)
aiteamforge setup # Enable "Fleet Monitor" as client
# Monitor entire fleet
open http://localhost:3000- macOS Big Sur or later
- Homebrew package manager
- Python 3 (3.8 or later)
- Node.js (18.0 or later)
- jq JSON processor
- Git version control
- GitHub CLI (
gh) - iTerm2 terminal emulator
- Claude Code AI pair programmer
The setup wizard will check for and offer to install missing dependencies.
- Tailscale - For multi-machine coordination
- ImageMagick - For avatar/image processing
- tmux - For Fleet Monitor
Framework (Homebrew-managed):
$(brew --prefix)/opt/aiteamforge/libexec/
├── bin/ # Core executables
├── scripts/ # Automation scripts
├── config/templates/ # Configuration templates
├── docs/ # Documentation
├── skills/ # Claude Code skills
├── lcars-ui/ # LCARS Kanban UI
├── kanban-hooks/ # Kanban automation
└── ...
Working Directory (user-managed):
~/aiteamforge/ # Default location
├── templates/ # Copied from framework
├── docs/ # Copied from framework
├── skills/ # Copied from framework
├── kanban/ # Kanban board data
├── teams/ # Team configurations
├── scripts/ # Generated scripts
└── .aiteamforge-config # Installation metadata
-
Framework Layer (
$(brew --prefix)/opt/aiteamforge/libexec/)- Installed via Homebrew
- Read-only template files
- Upgraded via
brew upgrade aiteamforge
-
Working Layer (
~/aiteamforgeor custom location)- Created by
aiteamforge setup - User-specific configuration
- Kanban data, team configs, generated scripts
- Preserved across framework upgrades
- Created by
Web-based kanban board with Star Trek LCARS interface:
- Real-time board updates
- Multi-team support
- Agent status tracking
- Calendar integration
- Health monitoring
Pre-configured teams with personas:
- Academy - Infrastructure and tooling
- iOS - iOS app development
- Android - Android app development
- Firebase - Backend/cloud functions
- Command - Strategic planning
- DNS - DNS framework
- Freelance - Client projects
- Legal - Legal/compliance
- MainEvent - Cross-platform coordination
- Medical - Health/diagnostics
AI agents with team-specific personas:
- Automated kanban tracking
- Session management
- Tool use hooks
- Custom skills and workflows
Multi-machine coordination (optional):
- Monitor multiple aiteamforge installations
- Centralized kanban aggregation
- Cross-machine agent status
- Tailscale integration
aiteamforge setup --install-dir ~/my-custom-location# Upgrade framework
brew upgrade aiteamforge
# Upgrade working directory
aiteamforge setup --upgradeWhy this exists: When AITeamForge installs its iTerm2 dynamic profiles for the
first time, the installer skips the file on subsequent runs to protect your
customizations. This means bug fixes shipped in the profile (such as the v0.8.27
Mouse Reporting and Dynamic Profile Parent Name corrections) do not propagate
automatically when you run brew upgrade aiteamforge.
When to use it: After brew upgrade aiteamforge when the release notes mention
dynamic profile changes — for example, a fix to click-to-navigate in the LCARS Web
browser tab, or a correction to profile inheritance.
How to run it:
aiteamforge setup --refresh-profilesWhat gets refreshed (AITeamForge-owned keys):
Name— profile identity used by iTerm2 for lookupsGuid— immutable profile anchorTags— AITeamForge visibility tag (["aiteamforge"])Custom Command— must be"Browser"for LCARS Web inline browser modeMouse Reporting— required for click-to-navigate in the browser tabDynamic Profile Parent Name— links the profile to its parent for font/color inheritanceBackground Color— functional pure-black background behind the browser viewInitial URL— source placeholder URL (overwritten at team start with the correct port)
What is preserved (your settings):
- Fonts (
Normal Font,Non Ascii Font, and related rendering flags) - Colors (
Foreground Colorand any other color customizations) - Window geometry (
Columns,Rows,Window Type) - Transparency and blur settings
- Scrollback buffer size
- Key mappings and triggers
- All other personal preferences
Rollback: Before writing changes the merge tool creates a timestamped backup of your existing profiles file at the same location:
~/Library/Application Support/iTerm2/DynamicProfiles/aiteamforge-lcars.json.bak-<timestamp>
Restore it by copying it back over the live file and letting iTerm2 hot-reload.
For maintainers who want to know which keys are in which bucket and why, see docs/refresh-profiles-design.md.
The cockpit profile is a minimal install for machines that connect into teams running on a primary dev machine over Tailscale — without running any team infrastructure locally.
Who it is for: Laptop users, second workstations, or thin clients that need to drive a remote AITeamForge session without maintaining local kanban boards, LCARS servers, or team directories.
How to install:
aiteamforge setup --cockpit-only
# --connect-only is an accepted aliasThe setup wizard skips team selection entirely — all connect scripts are installed unconditionally.
What a cockpit install includes:
| Component | Included |
|---|---|
<team>-connect.sh scripts (all teams) |
Yes |
Python venv with iterm2 package |
Yes |
iterm2_window_manager.py |
Yes |
| LCARS dynamic profile JSON | Yes |
set-lcars-profile-browser.py (iTerm2 plugin) |
Yes |
What a cockpit install excludes:
| Component | Excluded |
|---|---|
Team working directories (~/aiteamforge/<team>/) |
Omitted |
Kanban boards and kb-* helpers |
Omitted |
| LCARS UI server | Omitted |
| Persona markdown files and agent avatars | Omitted |
| Shell aliases, cc-aliases, statusline scripts | Omitted |
| Team zshrc fragments | Omitted |
Typical workflow:
# On cockpit machine: connect into a remote team running on another host
academy-connect.sh darren-m4-mini
# The remote team's full environment opens in iTerm2 over Tailscale
# Your local machine stays clean — no team state is created or maintained hereSwitching to a full install: There is no in-place migration from cockpit to full.
Re-run aiteamforge setup (without --cockpit-only) on the machine to perform a
complete installation.
Health checks: aiteamforge doctor recognizes the cockpit profile via a
.install-profile marker file and does not report missing kanban boards, LCARS
servers, or team directories as errors.
# Remove configuration (keeps framework)
aiteamforge setup --uninstall
# Remove framework
brew uninstall aiteamforgeaiteamforge doctorChecks:
- External dependencies
- Framework installation
- Configuration files
- Running services
- File permissions
aiteamforge doctor --verboseaiteamforge doctor --check dependencies
aiteamforge doctor --check services
aiteamforge doctor --check configLCARS server not starting
# Check port 8082 availability
lsof -i :8082
# Start manually
cd ~/aiteamforge/lcars-ui
python3 server.pyClaude Code not authenticated
claude auth loginGitHub CLI not authenticated
gh auth loginMissing dependencies
aiteamforge doctor
# Follow install instructions for missing depsComprehensive documentation is available in the docs/ directory:
- Quick Start - Get up and running in 5 minutes
- Installation Guide - Complete installation instructions
- User Guide - Day-to-day usage and commands
- Multi-Machine Setup - Fleet Monitor and multi-machine coordination
- Architecture - Technical architecture and design
- Team Reference - Complete team and agent reference
- Troubleshooting - Problem solving guide
- Contributing - How to contribute to aiteamforge
After installation, docs are also available locally:
ls ~/aiteamforge/docs/
cat ~/aiteamforge/docs/QUICK_START.mdAITeamForge is a comprehensive macOS development environment that combines:
- AI pair programming with Claude Code agents
- Visual task management with LCARS Kanban
- Multi-machine coordination with Fleet Monitor
- Terminal automation with iTerm2 integration
No, it's optional. AITeamForge works standalone, but Claude Code integration provides AI-assisted development with automatic kanban tracking.
Yes! AITeamForge supports both solo and team development. Fleet Monitor enables multi-machine coordination perfect for distributed teams.
AITeamForge includes 10 pre-configured teams:
- Platform: iOS, Android, Firebase
- Infrastructure: Academy, DNS Framework
- Project-Based: Freelance
- Coordination: MainEvent
- Strategic: Command, Legal, Medical
See Team Reference for details.
Yes! Teams are defined in data files, so you can add custom teams without modifying code. See ADDING_A_TEAM.
- Framework: ~100 MB
- Working directory: ~500 MB (excluding your kanban data and worktrees)
- Full installation: 2-5 GB depending on how many teams you install
Yes! The two-layer architecture separates the framework (Homebrew-managed) from your working directory (user-managed). Upgrades never touch your kanban data or configurations.
aiteamforge uninstall # Remove configuration
brew uninstall aiteamforge
brew untap DoubleNode/aiteamforgeYour ~/aiteamforge/ directory is preserved. Delete it manually if desired.
Run aiteamforge doctor for diagnostics. See Troubleshooting for common issues and solutions.
# Clone this tap
brew tap DoubleNode/aiteamforge
cd $(brew --repository DoubleNode/aiteamforge)
# Edit formula
vim Formula/aiteamforge.rb
# Test formula
brew install --build-from-source aiteamforge
brew test aiteamforge# Run formula tests
brew test aiteamforge
# Manual testing
aiteamforge doctor --verboseMIT License - See formula for details
- Issues: https://github.com/DoubleNode/aiteamforge/issues
- Documentation:
~/aiteamforge/docs/ - Health Check:
aiteamforge doctor
Current version: 1.0.0
Development Status: Phase 3 Complete (Interactive Setup Wizard)
- ✅ Setup wizard implemented
- ✅ LCARS-styled UI library
- 🚧 Installer modules (Phases 4-8 in progress)
aiteamforge --version