| name | scaffold |
| description | Creates a new project from templates with full setup including directory structure, dependency management, git initialization, CI/CD, testing, and documentation. Supports Python, Node.js, Rust, Go, and 14 AI agent framework templates. Use when the user wants to start a new project, bootstrap a codebase, or create a project from a template. |
| allowed-tools | ["Bash","Read","Edit","Grep","Glob"] |
| disable-model-invocation | true |
| user-invocable | true |
| argument-hint | <template-name> [project-name] |
Create a new project using the specified template: $ARGUMENTS
Output Format: Use clear section headers, timestamps for each step, and progress indicators for setup operations.
Tool Validation: First verify that required tools are installed (git, uv, gh) and the target directory is valid.
Python Project Defaults: All Python projects use UV for virtual environment management and package installation. Never use pip, pipenv, or conda.
Template Options:
python-lib - Python library with pyproject.toml, tests, CI/CDpython-cli - Python CLI application with Click/Typerpython-web - FastAPI web applicationpython-data - Data science project with Jupyter notebookspython-langchain - LangChain agent project with LangGraph supportpython-adk - Google Agent Development Kit multi-agent systempython-crewai - CrewAI multi-agent orchestration frameworkpython-autogen - Microsoft AutoGen conversational agentspython-swarm - OpenAI Swarm lightweight agent frameworkpython-phidata - Phidata AI assistant frameworkpython-llama-index - LlamaIndex RAG agents with document indexingpython-haystack - Deepset Haystack NLP pipelinespython-semantic-kernel - Microsoft Semantic Kernel plugin architecturepython-agency-swarm - Agency Swarm hierarchical agent frameworknode-lib - Node.js library with TypeScriptnode-web - React/Next.js web applicationrust-lib - Rust library with Cargo.tomlgo-cli - Go CLI applicationcustom - Interactive template selection
Template Loading: Read the template definition from ${CLAUDE_SKILL_DIR}/templates/<template-name>.md for the selected template's directory structure and dependencies.
Project Setup Steps:
- Template Selection: Determine template from argument or prompt interactively
- Load Template: Read the template file from
${CLAUDE_SKILL_DIR}/templates/<template-name>.md
- Directory Creation: Create project directory with proper naming conventions
- Template Scaffolding: Create the directory structure and files defined in the template
- Dependency Management:
- For Python: Use
uv init and uv add exclusively for all dependencies
- For Python: Create virtual environment with
uv venv and activate with uv run
- For other languages: Install language-specific dependencies
- Git Initialization: Initialize repo, create .gitignore, initial commit
- CI/CD Setup: Create GitHub Actions workflows, testing and linting pipelines
- Development Environment: Setup IDE config files, development scripts
- Documentation: Generate README, CONTRIBUTING.md, documentation framework
- Testing Framework: Setup test directory, sample tests, configure runners
- Quality Tools: Configure linting, formatters, type checking
- Final Validation:
- For Python: Run
uv run python -m pytest and uv build
- For other languages: Run initial build/compile and basic tests
Interactive Prompts:
- Project name (default: directory name)
- Author name and email
- License type (MIT, Apache, etc.)
- Description
- Python version (for Python projects)
Template Variables:
{{PROJECT_NAME}} - Project name{{AUTHOR_NAME}} - Author name{{AUTHOR_EMAIL}} - Author email{{DESCRIPTION}} - Project description{{LICENSE}} - License type{{YEAR}} - Current year{{PYTHON_VERSION}} - Python version
Error Handling: If any step fails, attempt to fix the issue automatically. If unable to fix, report the specific error and provide cleanup instructions.
Post-Setup Instructions:
- Display next steps for development
- Show available commands (test, build, etc.)
- Provide quick start guide
- Suggest development workflow
