A production-ready repository and CLI tool for managing reusable Claude Code and Roo Code configurations across multiple engineering domains.
- 🎯 5 Engineering Domains: DevOps, Python Backend, JavaScript/TypeScript, Data Engineering, Machine Learning
- 🤖 5 Agent Roles per Domain: Researcher, Planner, Architect, Implementer, Reviewer
- 🔧 Multi-Platform Support: Works with both Roo Code and Claude Code
- 📦 Easy Installation: Simple CLI tool with one-command setup
- 🔄 Flexible Sourcing: Pull from public repo, custom repo, or local path
- 🚀 Production-Ready: Comprehensive configurations with best practices
Linux/macOS:
curl -sSL https://raw.githubusercontent.com/username/hailow/main/scripts/install.sh | bashOr download binary directly from GitHub Releases
# Install single domain for Roo Code
hailow install devops-engineer --platform roo
# Install multiple domains
hailow install devops-engineer python-backend-engineer
# Install all domains
hailow install --all
# Install for Claude Code
hailow install python-backend-engineer --platform claude| Domain | Description | Aliases |
|---|---|---|
devops-engineer |
Infrastructure, CI/CD, deployment, operations | devops, ops |
python-backend-engineer |
Python backend development, APIs, services | python, py |
js-ts-software-engineer |
JavaScript/TypeScript development | jsts, js, ts |
data-engineer |
Data pipelines, ETL, data quality | data, de |
machine-learning-engineer |
ML models, training, deployment | mleng, ml |
Each domain includes:
-
Agents (5 specialized prompts):
researcher- Investigates requirements and contextplanner- Breaks down work into actionable tasksarchitect- Designs technical solutionsimplementer- Executes implementationreviewer- Validates quality and correctness
-
Rules: Domain-specific coding standards and best practices
-
Skills: Recommended skills to install for the domain
-
Commands: Workflow commands and shortcuts
-
Contexts: Placeholder for user-specific project context
All domains follow a structured 5-phase workflow:
Research → Plan → Architect → Implement → Review
See AGENT_WORKFLOW.md for details.
# Install domains
hailow install <domain>... # Install specific domains
hailow install --all # Install all domains
hailow install --dry-run # Preview installation
# List information
hailow list domains # List available domains
hailow list platforms # List supported platforms
hailow list installed # Show installed domains
# Manage installations
hailow remove <domain>... # Remove domains
hailow remove --all # Remove all domains
hailow upgrade [domain]... # Upgrade to latest versions
# Configuration
hailow config show # Show current config
hailow config set <key> <value> # Set config value
# Version
hailow version # Show version infoThe CLI uses ~/.hailow/config.yaml:
# Default source
source:
type: git
url: https://github.com/Harvey-N-Lab/hailow
branch: main
# Default platform
platform: roo
# Installation behavior
install:
on_conflict: skip # skip, overwrite, prompt
include_general: true
backup: true# Configure
hailow config set source.url https://github.com/mycompany/configs
# Or use flag
hailow install devops-engineer --source https://github.com/mycompany/configs# Configure
hailow config set source.type local
hailow config set source.path /path/to/configs
# Or use flag
hailow install python-backend-engineer --source /path/to/configs# Install DevOps domain
hailow install devops-engineer --platform roo
# Your workspace now has:
# - .agents/devops-researcher.md
# - .agents/devops-planner.md
# - .agents/devops-architect.md
# - .agents/devops-implementer.md
# - .agents/devops-reviewer.md
# - .agents/rules/ (infrastructure, security, deployment standards)
# - .agents/skills/ (terraform-helper, kubernetes-assistant, etc.)
# - .agents/commands/ (deploy, rollback, health-check)
# - AGENT_WORKFLOW.md (workflow guidance)# Install frontend and backend domains
hailow install js-ts-software-engineer python-backend-engineer devops-engineer
# Team now has configurations for:
# - Frontend development (React, TypeScript, etc.)
# - Backend development (Python, API design, etc.)
# - DevOps operations (deployment, infrastructure, etc.)# Install data and ML domains
hailow install data-engineer machine-learning-engineer --all
# Team has configurations for:
# - Data pipelines and ETL
# - ML model development and deployment.
├── general/ # Cross-domain rules
├── devops-engineer/ # DevOps domain
├── python-backend-engineer/ # Python backend domain
├── js-ts-software-engineer/ # JS/TS domain
├── data-engineer/ # Data engineering domain
├── machine-learning-engineer/ # ML engineering domain
├── domains/ # Domain metadata
├── cmd/hailow/ # CLI source code
├── internal/ # Internal packages
└── .github/workflows/ # CI/CD
See REPOSITORY_STRUCTURE.md for details.
- INSTALL.md - Detailed installation instructions
- UPGRADE.md - Upgrade guide
- AGENT_WORKFLOW.md - Agent collaboration workflow
- ARCHITECTURE.md - Architecture decisions
- CLI_DESIGN.md - CLI design and usage
- Domain Documentation:
- DevOps Engineer Rules
- Python Backend Rules
- And more...
# Clone repository
git clone https://github.com/Harvey-N-Lab/hailow
cd hailow
# Build CLI
make build
# Run tests
make test
# Install locally
make install- Go 1.21 or later
- Git (for fetching remote configs)
Contributions are welcome! Please see CONTRIBUTING.md for guidelines.
- Create domain directory:
<domain-name>/ - Add required subdirectories:
agents/,rules/,skills/,commands/,contexts/ - Create 5 agent files with domain-specific prompts
- Add domain to
domains/domains.yaml - Update documentation
- Submit pull request
- Shell completion (bash, zsh, fish)
- Interactive domain selection
- Domain templates for creating custom domains
- Validation of custom domains
- Diff view for upgrades
- Backup/restore functionality
- Web UI for browsing domains
MIT License - see LICENSE for details.
- Issues: GitHub Issues
- Discussions: GitHub Discussions
- Documentation: Wiki
Built for engineers who want to level up their AI-assisted development workflow.
Made with ❤️ for the developer community