claude-setup

My full Claude Code configuration -- 118 agents, 49 skills, 148 slash commands, and the GSD (Get Shit Done) workflow engine that ties it all together.

This is the exact setup I use daily for building software with Claude Code. Clone it, run the installer, and you're up.

Quick Install

One-Liner (send this to Claude)

Open Claude Code and paste:

Claude will handle the rest.

Manual Install

• Backs up your existing ~/.claude config
• Copies all agents, skills, commands, hooks, and the GSD engine
• Installs settings.json and CLAUDE.md
• Sets up the claude-flow MCP server

Key Workflows

Auto-Triggered Skills

These fire automatically based on context -- you don't need to call them:

• About to build something? - brainstorming activates first
• Writing code? - test-driven-development writes tests first
• Hit a bug? - systematic-debugging does 4-phase root cause analysis
• Claiming done? - verification-before-completion demands evidence
• Finished a feature? - requesting-code-review runs quality gates

Model Tiering (save money)

The config routes work to the right model tier:

Sonnet does the work. Opus checks the work. This cuts costs significantly.

Optional: Telegram Notifications

Get notified on your phone when Claude finishes tasks:

Prerequisites

• Claude Code CLI (npm install -g @anthropic-ai/claude-code)
• Node.js 18+
• Git