Everything you need to go from zero to your first project. A step-by-step guide covering installation, core commands, and best practices.
Follow this guided path from fundamentals to your first deployment. Each stage builds on the previous one.
Learn what OpenClaw is, its architecture, and why teams choose it for modern application pipelines.
~15 min readSet up OpenClaw on your machine, configure your environment, and verify everything works.
~10 min setupFollow a hands-on tutorial to scaffold, develop, and run a working project locally.
~25 min tutorialShip your project to production, set up CI/CD hooks, and learn the deployment lifecycle.
~20 min guideOpenClaw works on macOS, Linux, and Windows. Choose your preferred method below.
Recommended for most users
One-line install without Node.js
Containerised development environment
Create a working task tracker from scratch to learn OpenClaw's core workflow.
Generate a new project with the task-tracker template. OpenClaw creates the folder structure, config files, and installs dependencies automatically.
Open schema/tasks.claw and define the Task entity. OpenClaw schemas are declarative — just describe the shape of your data.
Use the generate command to create CRUD endpoints from your schema. OpenClaw reads the entity and creates type-safe handlers.
Launch the development server with hot reload. OpenClaw watches for file changes and restarts automatically.
Run the built-in test suite to make sure everything works. OpenClaw auto-generates smoke tests for your routes.
A quick-reference cheat sheet for the most essential OpenClaw CLI commands.
openclaw init <name>
Create a new project with default configuration. Add --template to use a specific starter.
openclaw dev
Start local dev server with hot reload, file watching, and the built-in dashboard.
openclaw generate <type>
Auto-generate routes, handlers, tests, or migrations from your schema definitions.
openclaw test
Run all tests. Use --watch for live re-runs or --coverage for a coverage report.
openclaw deploy
Build and deploy to your configured target (cloud, container, or edge). Runs checks automatically.
openclaw doctor
Check your environment: Node version, dependencies, config validity, and port availability.
openclaw migrate
Apply pending data migrations. Use --dry-run to preview changes without executing.
openclaw plugin add <name>
Install community or official plugins. Use plugin list to browse available extensions.
Stuck? These are the problems beginners run into most often, with proven solutions.
openclaw doctor --path to print the correct PATH entry, then add it to your ~/.bashrc or ~/.zshrc and restart your terminal.openclaw dev --port 4001 to use a different port. You can also set port: 4001 in openclaw.config.yaml..claw file for typos in type names. Common mistakes: using string instead of String, or missing the closing brace. Run openclaw lint schema/ for detailed error locations.echo 65536 | sudo tee /proc/sys/fs/inotify/max_user_watches. On macOS, this usually means the fsevents module is missing — reinstall dependencies.openclaw build before deploying. If the build itself fails, check that all environment variables referenced in your config are defined. Use openclaw env check to list missing variables.node: '20' to your CI matrix and make sure openclaw doctor runs as the first CI step.You've got the basics down. Here's the recommended path to level up your skills.
Extend OpenClaw with community plugins for auth, database connectors, caching, and more. Learn to write your own plugins.
IntermediateDefine one-to-many and many-to-many relationships between entities. Manage data evolution with versioned migrations.
IntermediateGo beyond auto-generated smoke tests. Write integration tests, mock external services, and measure coverage thresholds.
IntermediateLearn blue-green deployments, environment-specific configs, secrets management, and rollback procedures.
AdvancedJoin the open-source community. Report issues, submit pull requests, write plugins, and help improve documentation.
Community