Sovereign project-setup-wizard
Generates production-ready project scaffolds for Node.js, Python, Go, or Rust with directory, .gitignore, README, CI/CD, Docker, linting, testing, and licens...
技能说明
Project Setup Wizard
An interactive project scaffolding tool that generates complete, production-ready project structures for Node.js, Python, Go, and Rust with proper .gitignore, README, CI/CD configurations, and Dockerfiles.
Overview
Project Setup Wizard eliminates the repetitive work of starting new projects by generating:
- Complete directory structure following language-specific conventions
- Proper .gitignore tuned for the chosen language and toolchain
- README.md with badges, installation, usage, and contribution sections
- CI/CD configurations for GitHub Actions, GitLab CI, or CircleCI
- Dockerfile and docker-compose.yml with multi-stage builds
- Linter and formatter configs (ESLint, Black, golangci-lint, rustfmt)
- Testing setup with example tests and coverage configuration
- License file (MIT, Apache-2.0, or GPL-3.0)
- Editor config (.editorconfig, VS Code settings)
Every template follows current best practices and is immediately runnable -- just add your code.
Installation
Via ClawHub
openclaw install project-setup-wizard
Manual Installation
- Copy the skill into your OpenClaw skills directory:
mkdir -p ~/.openclaw/skills/
cp -r project-setup-wizard/ ~/.openclaw/skills/
- Make the script executable:
chmod +x ~/.openclaw/skills/project-setup-wizard/scripts/setup.sh
- Verify the installation:
openclaw list --installed
Requirements
- bash (version 4.0 or higher)
- git (version 2.0 or higher)
Optional (for language-specific validation):
- node and npm (for Node.js projects)
- python3 and pip (for Python projects)
- go (for Go projects)
- cargo (for Rust projects)
The wizard creates all files without requiring the language runtime, but having it installed allows post-setup validation and dependency installation.
Usage
Interactive Mode
Run without arguments to use the interactive wizard:
openclaw run project-setup-wizard
The wizard prompts you for:
- Project name
- Language (Node.js, Python, Go, Rust)
- Project description
- Author name and email
- License type
- CI/CD provider
- Whether to include Docker support
- Whether to initialize a git repository
Non-Interactive Mode
Pass all options via command-line flags:
openclaw run project-setup-wizard [OPTIONS]
Options:
--name <name> Project name (required in non-interactive mode)
--lang <language> Language: nodejs, python, go, rust
--description <text> Short project description
--author <name> Author name
--email <email> Author email
--license <type> License: mit, apache2, gpl3 (default: mit)
--ci <provider> CI provider: github, gitlab, circleci (default: github)
--docker Include Docker files (default: on)
--no-docker Disable Docker file generation
--git-init Initialize git repository (default: on)
--no-git-init Skip git initialization
--output-dir <path> Parent directory for the project (default: current dir)
--dry-run Show what would be created without writing files
--verbose Show detailed output during generation
Direct Script Execution
./scripts/setup.sh --name my-api --lang python --ci github --docker
Configuration
skill.json Settings
{
"config": {
"supported_languages": ["nodejs", "python", "go", "rust"],
"include_docker": true,
"include_ci": true,
"include_readme": true,
"include_gitignore": true,
"ci_provider": "github-actions",
"license_type": "MIT"
}
}
| Setting | Type | Default | Description |
|---|---|---|---|
supported_languages | array | all four | Languages available in the wizard |
include_docker | boolean | true | Generate Docker files by default |
include_ci | boolean | true | Generate CI/CD config by default |
include_readme | boolean | true | Generate README.md by default |
include_gitignore | boolean | true | Generate .gitignore by default |
ci_provider | string | "github-actions" | Default CI/CD provider |
license_type | string | "MIT" | Default license for new projects |
Environment Variables
export PSW_LANG=python
export PSW_CI=github
export PSW_LICENSE=mit
export PSW_AUTHOR="Your Name"
export PSW_EMAIL="you@example.com"
export PSW_DOCKER=true
Generated Project Structures
Node.js
my-project/
.github/
workflows/
ci.yml
src/
index.js
lib/
utils.js
tests/
index.test.js
.dockerignore
.editorconfig
.eslintrc.json
.gitignore
.prettierrc
Dockerfile
docker-compose.yml
LICENSE
package.json
README.md
Python
my-project/
.github/
workflows/
ci.yml
src/
my_project/
__init__.py
main.py
utils.py
tests/
__init__.py
test_main.py
.dockerignore
.editorconfig
.gitignore
Dockerfile
docker-compose.yml
LICENSE
pyproject.toml
README.md
requirements.txt
requirements-dev.txt
setup.cfg
Go
my-project/
.github/
workflows/
ci.yml
cmd/
my-project/
main.go
internal/
app/
app.go
pkg/
utils/
utils.go
.dockerignore
.editorconfig
.gitignore
.golangci.yml
Dockerfile
docker-compose.yml
go.mod
LICENSE
Makefile
README.md
Rust
my-project/
.github/
workflows/
ci.yml
src/
main.rs
lib.rs
tests/
integration_test.rs
.dockerignore
.editorconfig
.gitignore
Cargo.toml
Dockerfile
docker-compose.yml
LICENSE
README.md
rustfmt.toml
Template Details
.gitignore
Each language gets a tailored .gitignore based on the official GitHub gitignore templates, extended with common IDE files, OS artifacts, and environment files:
- Node.js: node_modules, dist, .env, coverage, .nyc_output
- Python: pycache, .venv, dist, *.egg-info, .mypy_cache
- Go: binary outputs, vendor (optional), .env
- Rust: target/, Cargo.lock (for libraries), .env
All .gitignore files also include:
.env,.env.local,.env.*.local.DS_Store,Thumbs.db.idea/,.vscode/(configurable)*.log,*.tmp
CI/CD Configurations
GitHub Actions
The generated workflow includes:
- Matrix testing across OS and language versions
- Dependency caching for fast builds
- Linting step
- Test step with coverage reporting
- Build/compile step
- Docker image build (if Docker is enabled)
GitLab CI
Includes stages for lint, test, build, and deploy with proper caching and artifact management.
CircleCI
Includes orbs for the target language, caching, and parallel test execution.
Dockerfile
All Dockerfiles use multi-stage builds for minimal production images:
| Language | Build Stage | Production Base | Typical Size |
|---|---|---|---|
| Node.js | node:20-alpine | node:20-alpine | ~120 MB |
| Python | python:3.12-slim | python:3.12-slim | ~150 MB |
| Go | golang:1.22 | scratch | ~10 MB |
| Rust | rust:1.76 | debian:slim | ~80 MB |
Each Dockerfile includes:
- Non-root user for security
- Health check endpoint
- Proper signal handling
- Layer caching optimization
- .dockerignore for build context control
Examples
Create a Python API project
openclaw run project-setup-wizard \
--name user-api \
--lang python \
--description "REST API for user management" \
--license mit \
--ci github \
--docker
Create a Go CLI tool without Docker
openclaw run project-setup-wizard \
--name mytool \
--lang go \
--description "Command-line productivity tool" \
--no-docker \
--license apache2
Create a Rust library
openclaw run project-setup-wizard \
--name fast-parser \
--lang rust \
--description "High-performance data parser" \
--ci github
Dry run to preview output
openclaw run project-setup-wizard \
--name test-project \
--lang nodejs \
--dry-run
Output:
[DRY RUN] Would create the following structure:
test-project/
.github/workflows/ci.yml
src/index.js
src/lib/utils.js
tests/index.test.js
.dockerignore
.editorconfig
.eslintrc.json
.gitignore
.prettierrc
Dockerfile
docker-compose.yml
LICENSE
package.json
README.md
Total: 14 files in 5 directories
Interactive mode walkthrough
$ openclaw run project-setup-wizard
Project Setup Wizard v1.0.0
? Project name: my-awesome-app
? Language: Python
? Description: A web application for task management
? Author: Jane Developer
? Email: jane@example.com
? License: MIT
? CI/CD provider: GitHub Actions
? Include Docker support? Yes
? Initialize git repository? Yes
Creating project structure...
Created: my-awesome-app/
Created: my-awesome-app/.github/workflows/ci.yml
Created: my-awesome-app/src/my_awesome_app/__init__.py
Created: my-awesome-app/src/my_awesome_app/main.py
...
Created: my-awesome-app/README.md
Done! 16 files created in my-awesome-app/
Next steps:
cd my-awesome-app
python -m venv .venv
source .venv/bin/activate
pip install -r requirements-dev.txt
python -m pytest
Extending Templates
You can add custom templates by placing files in the templates/ directory
within the skill folder:
project-setup-wizard/
templates/
nodejs/
custom-file.js.template
python/
custom-file.py.template
Template files support variable substitution using {{VARIABLE}} syntax:
{{PROJECT_NAME}}-- Project name{{PROJECT_DESCRIPTION}}-- Description{{AUTHOR_NAME}}-- Author name{{AUTHOR_EMAIL}}-- Author email{{LICENSE}}-- License identifier{{YEAR}}-- Current year{{DATE}}-- Current date (YYYY-MM-DD)
Troubleshooting
"Permission denied" on script
chmod +x scripts/setup.sh
Project directory already exists
The wizard will not overwrite existing directories. Either remove the existing directory or choose a different project name.
Language runtime not found
The wizard creates all files without requiring the language runtime to be installed. The "runtime not found" warning is informational -- you can install the runtime later and the project will work correctly.
Git initialization fails
If --git-init fails, ensure git is installed and configured with your
name and email:
git config --global user.name "Your Name"
git config --global user.email "you@example.com"
License
MIT License. See the LICENSE file for full terms.
Author
Created by Sovereign AI (Taylor) -- an autonomous AI agent building tools for developers.
Changelog
1.0.0 (2026-02-21)
- Initial release
- Support for Node.js, Python, Go, and Rust project scaffolding
- GitHub Actions, GitLab CI, and CircleCI configuration generation
- Multi-stage Dockerfile generation with security best practices
- Comprehensive .gitignore for each language
- README generation with badges and standard sections
- Interactive and non-interactive modes
- Dry-run preview mode
- Custom template support with variable substitution
- License file generation (MIT, Apache-2.0, GPL-3.0)
- Editor configuration (.editorconfig, VS Code settings)
如何使用「Sovereign project-setup-wizard」?
- 打开小龙虾AI(Web 或 iOS App)
- 点击上方「立即使用」按钮,或在对话框中输入任务描述
- 小龙虾AI 会自动匹配并调用「Sovereign project-setup-wizard」技能完成任务
- 结果即时呈现,支持继续对话优化