Skip to main content

Crack the egg.
Hatch better agents.

One command installs a battle-tested agentic coding setup into any repository — 11 agents, 22 skills, 18 rules, 25 commands, and MCP integrations for all major AI coding tools.

Get StartedCore Concepts
11
Agents
22
Skills
18
Rules
25
Commands
hatch3r logo

Quick start

Get up and running in seconds

1

Install hatch3r

Run the init command in your repository:
npx hatch3r init
hatch3r will detect your repository and ask which AI coding tools you use.
2

Select your tools

Choose from supported tools:
  • Cursor — .mdc rules, agents, skills, commands, MCP config
  • GitHub Copilot — instructions, prompts, GitHub agents
  • Claude Code — CLAUDE.md, skills, .mcp.json
  • OpenCode — AGENTS.md, opencode.json
  • Windsurf — .windsurfrules
  • Amp, Codex CLI, Gemini CLI, Cline — Full support
You can select multiple tools — hatch3r generates native configuration for each
3

Start using agents

Your repository now includes:
  • 11 specialized agents (reviewer, test-writer, security-auditor, implementer, etc.)
  • 22 skills (feature, bug-fix, refactor, release, incident-response, etc.)
  • 18 rules (code standards, API design, testing, security patterns, etc.)
  • 25 commands (board management, planning, quality audits, workflows)
  • MCP integrations for GitHub, Context7, Filesystem, Playwright, Brave Search
Run any command in your AI coding tool — for example: board-init, feature-plan, or healthcheck

Key features

Everything you need for production-ready AI-assisted development

11 Specialized Agents

Code reviewer, test writer, security auditor, implementer, researcher, and more — each with a specific role and expertise.

22 Skills

Feature development, bug fixing, refactoring, releases, incident response, architecture review, and more.

18 Rules

Code standards, API design, testing patterns, security guidelines, observability, and workflow conventions.

25 Commands

Board management, planning workflows, quality audits, security scans, release automation, and utilities.

Multi-Tool Support

Native adapters for Cursor, GitHub Copilot, Claude Code, OpenCode, Windsurf, Amp, Codex, Gemini, and Cline.

MCP Integrations

Built-in MCP servers for GitHub, Context7, Filesystem, Playwright, and Brave Search with automatic configuration.

Sub-Agentic Architecture

Parallel delegation and dependency-aware orchestration for complex multi-step workflows.

Board Management

GitHub Projects V2 integration with automated issue creation, status tracking, and dependency analysis.

How it works

One canonical source, multiple tool outputs

hatch3r maintains one source of truth in .agents/ and generates native configuration for each tool you use. Run hatch3r sync to regenerate after changes.

Complete workflow

From initialization to release

New Projects (Greenfield)

Start from scratch with project-spec to generate documentation, then roadmap to create a phased plan.

Learn more

Existing Projects (Brownfield)

Add hatch3r to existing codebases with codebase-map to analyze what’s there and plan improvements.

Learn more

Resources

Guides and references

CLI Reference

Complete reference for all CLI commands: init, sync, update, validate, status

Agent Commands

25+ commands for board management, planning, quality audits, and workflows

Customization

Customize agents, skills, rules, and commands for your project

MCP Setup

Configure MCP servers and manage API keys securely

Model Selection

Configure AI models per agent for optimal performance

Troubleshooting

Common issues and solutions

Ready to hatch better agents?

Install hatch3r in your repository and get a production-ready agentic coding setup in seconds.

Get Started Now