🇰🇷 한국어 버전
From a single line of an idea to a working CLI tool — the full journey with an AI agent
This tutorial is a hands-on vibe coding workbook documenting the actual process of building the vibe-ready-cli project with Claude Code. It contains real prompts, decisions, failures, and insights extracted directly from conversation logs.
- Developers new to Claude Code
- Anyone interested in vibe coding (AI agent-driven development)
- Those who want to apply harness engineering concepts to real projects
- Node.js >= 24
- Claude Code CLI installed (
npm install -g @anthropic-ai/claude-code) - Claude Code subscription (required for Agent SDK usage)
| Chapter | Title | Key Content |
|---|---|---|
| 01 | Ideation and Project Initialization | Writing the ideation document, generating CLAUDE.md with /init |
| 02 | Clarifying Requirements with Deep Interview | 10-round Q&A reducing ambiguity from 100% to 19% |
| 03 | MVP Implementation | Building analyzers, scorers, and reporters using Claude Agent SDK |
| 04 | Harness Engineering | Enhancing CLAUDE.md, AGENTS.md, settings.json, hook configuration |
| 05 | Contribution Guide and Auto-Validation Skill | Writing CONTRIBUTING.md, automated validation with the contribution-guard skill |
05:17 — Project started, ideation document written
05:25 — CLAUDE.md initially generated with /init
05:28 — Deep interview started (Round 1, ambiguity 100%)
05:49 — Deep interview completed (Round 10, ambiguity 19%)
05:50 — MVP implementation started (CLI + analysis engine + scorer + reporter)
06:30 — Harness engineering started
06:41 — Harness engineering committed
06:45 — Contribution guide + contribution-guard skill created
06:55 — Full commit and push completed
Approximately 1 hour 40 minutes (idea → working CLI + harness engineering + contribution framework)
last_generated: 2026-04-01T06:55:00Z last_log_session: ee266d92-9f23-48c9-9576-dc19f987f463