Quick Start

From install to your first AI-orchestrated feature in 5 minutes.

---

1. Install

curl -fsSL https://raw.githubusercontent.com/PlayIdea-Lab/cq/main/install.sh | bash

Verify:

cq --version
# cq v1.58 (darwin/arm64)

Linux / WSL2: The same one-liner works. cq is a single static binary with no runtime dependencies.

---

2. Start

cq claude    # Launches Claude Code with CQ tools

CQ registers 275+ MCP tools automatically. No manual .mcp.json editing required — cq init already handled it.

Other supported AIs:

cq cursor    # Cursor
cq codex     # OpenAI Codex CLI
cq gemini    # Google Gemini CLI

---

3. Your First /pi Session

Inside Claude Code, type:

/pi "add a health check endpoint to the API"

CQ will:

• Research the problem space and relevant codebase context
• Generate an idea document (.c4/ideas/health-check.md)
• Ask: "자동 구현" or "계획만"?

---

4. Auto-Implement

Choose "자동 구현" and CQ chains the full pipeline:

• /plan — EARS requirements → design → task breakdown
• /run — parallel workers execute tasks
• /finish — build → test → wire smoke → commit

No manual task creation. No context switching.

---

5. What Just Happened?

Artifact	Location
Idea doc	.c4/ideas/health-check.md
Spec	docs/specs/health-check.md
Tasks	.c4/tasks.db (C4 queue)
Code	Committed with conventional commits
Knowledge	Insights recorded for future sessions

CQ's external brain captures every decision so the next session starts smarter.

---

Works With

AI	Launch command
Claude Code	cq claude
ChatGPT	via mcp.pilab.kr (remote brain)
Cursor	cq cursor
Gemini CLI	cq gemini
Codex CLI	cq codex

---

Next Steps

• Dogfooding Workflow — full /pi → /finish guide
• Examples — step-by-step walkthroughs
• Wire Safety — 3-layer validation