init-project

Generate CLAUDE.md and supporting artifacts for new Claude Code projects.

Updated Mar 6, 2026
One-click install
npx skills add https://github.com/CopeeeTang/copeee-skills --skill init-project-copeeetang
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: init-project
Source: https://github.com/CopeeeTang/copeee-skills/tree/main/skills/init-project
Command: npx skills add https://github.com/CopeeeTang/copeee-skills --skill init-project-copeeetang

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

It quickly bootstraps a new Claude Code project by generating the essential guidance and research artifacts, so you start from a clean, well-structured baseline instead of starting blank and guessing what to write.

Core Features & Use Cases

  • Signal-driven project blueprinting: auto-detects language, package manager, repo/team/shared context, and project type (ML research vs web app) before asking questions.
  • Dry-run first, then write: previews exactly which files will be created and lets you edit/skip/cancel per file to prevent unwanted changes.
  • Minimal default artifacts for Claude Code: generates CLAUDE.md plus the right supporting files (research-notes.md and a project manifest like pyproject/package.json), with optional personal/team/public handling and safe gitignore additions.
  • ML research / Web app profiles: loads tailored question sets and reference templates for deeper setup (e.g., red-line constraints, data/compute/targets for ML).

Quick Start

Run init-project to bootstrap a new project blueprint with dry-run previews first, then approve the generated files.

Frequently Asked Questions about init-project

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I bootstrap a Claude Code project blueprint for a new ML research repo?

Bootstrapping a Claude Code project blueprint for ML research auto-detects your environment, asks targeted clarification questions about data and compute, and generates CLAUDE.md plus research-notes.md. It applies a dry-run first, previewing files for per-file approval before writing anything.

What is the best way to scaffold a web app project with CLAUDE.md and safe gitignore defaults?

Scaffolding a web app with safe gitignore defaults and CLAUDE.md is best handled by generating a project manifest like package.json alongside the core guidance file. The process uses signal-driven detection to identify your package manager and applies dry-run previews to prevent unwanted file writes.

Can I generate CLAUDE.md files without installing dependencies during project initialization?

Yes, you can generate CLAUDE.md without installing dependencies because this project initialization process intentionally defers dependency installation. It focuses strictly on outputting minimal default artifacts like the project manifest and research notes based on your approved dry-run preview.

Does the project blueprint generator support both CLI and data pipeline setups?

Yes, the project blueprint generator supports CLI and data pipeline setups alongside ML research and web apps. It loads tailored question sets and reference templates for each specific profile to ensure the generated CLAUDE.md and supporting artifacts match your project type.

Why use a dry-run preview before writing files when setting up a new project?

Using a dry-run preview before writing files prevents unwanted changes to your repository during project setup. It shows exactly which files will be created and allows you to edit, skip, or cancel on a per-file basis before the tool writes any final artifacts to disk.