GuideIntermediate
Git and Environments for n8n: Version, Test, and Ship Like a Pro
Ship your n8n workflows like a professional, not a hobbyist: version-controlled in Git, tested in a sandbox with synthetic data and test credentials, and promoted through dev, staging, and prod environments with rollback capability. By the end you produce exactly what the best offers in the market ask for in writing ("as version-controlled, documented JSON"), and you cross the line from "workflow builder" to "automation system owner." The default path throughout is self-hosted Community at zero cost; what only exists in Enterprise or Cloud is called out honestly, with the criteria for deciding when it's worth it.
- 48
- lessons
- 6
- modules
- English · Spanish
- available in
- Yes
- certificate
- Free
- access
Outcomes
What you'll be able to do
- Understand the difference between a workflow builder and a system owner, and the real limits of exporting and importing JSON by hand
- Learn Git from zero, applied to workflows: repository, staging and commits, reading a diff, branches, pushing to GitHub, and rollback
- Export workflows with the n8n CLI, separate credentials from code, normalize the JSON for clean diffs, and document for handoff
- Spin up isolated dev, staging, and prod environments with Docker Compose on top of the Self-Hosted AI Starter Kit v2, each with its own encryption key and credentials
- Test in a sandbox before production: test accounts and sandbox keys, synthetic data, dry runs, pinned data, and execution replay
- Test AI Agent node workflows at zero cost with local models, and verify results with the Evaluation node
- Promote workflows between environments, review changes as diffs (including AI-generated ones), and run a rollback with a runbook
- Set up a CI check in GitHub Actions that validates the workflow JSON on every commit
Before you start
What you need to bring
It's for you if...
- Anyone already building real n8n workflows who wants to stop handling versions by hand through exporting and importing JSON
- Freelancers and consultants who need to deliver "as version-controlled, documented JSON," exactly as market offers ask for it
- Teams who want a reproducible dev, staging, and prod flow on self-hosted Community, without paying for Enterprise
- Automators with no prior Git experience who want to learn it applied to their real use case
Requirements and materials
- Having built real workflows in n8n and being comfortable with the 2.0 editor
- Basic terminal skills (running commands, editing files) and Docker installed on your machine
- No prior Git experience required: it's taught from zero in Module 2
- Basic JavaScript helps in the Code node, but it's not mandatory
Content
The syllabus, module by module
Open any of them to see its lessons.
- 1. Introduction: from loose JSON to a repository
- 2. Workflow builder vs system owner
- 3. The limits of exporting and importing JSON
- 4. Anatomy of a workflow JSON
- 5. What changes and what breaks when you reimport
- 6. The mental model: workflow as code
- 7. Source control: Community vs Enterprise, with honesty
- 8. Project: audit and export a workflow
- 1. Introduction: Git applied to workflows
- 2. Install Git and create your first repository
- 3. Staging and commits: photos of your workflow
- 4. Reading a workflow's diff
- 5. Branches for changing without fear
- 6. Push to GitHub: backup and shared work
- 7. Rollback: going back to a version that worked
- 8. Project: a workflow under version control
- 1. Introduction: from the editor to a reproducible repository
- 2. Exporting with n8n's CLI
- 3. Separating credentials from workflows
- 4. Normalizing the JSON for clean diffs
- 5. Structuring the repository
- 6. Documenting a workflow for handoff
- 7. Automating export with a script
- 8. Project: from instance to documented repository
- 1. Introduction: what a real environment is
- 2. Reproducible base: the Self-Hosted AI Starter Kit
- 3. One Docker Compose per environment
- 4. Environment variables and encryption keys
- 5. Credentials per environment: test vs. production
- 6. External secrets and variables
- 7. Native environments (Enterprise): when it's worth paying
- 8. Project: three isolated environments running
- 1. Introduction: what testing in a sandbox means
- 2. Test accounts and sandbox API keys
- 3. Generating synthetic data
- 4. Dry runs and guarding side effects
- 5. Pinned data and execution replay
- 6. Testing AI Agent workflows at zero cost
- 7. Assertions with the Evaluation node
- 8. Project: a full sandbox test pass
- 1. Introduction: closing the delivery cycle
- 2. Promoting workflows between environments
- 3. Reviewing changes as diffs, AI included
- 4. MCP for building workflows against non-prod environments
- 5. Rollback and operations runbook
- 6. Native Git (Enterprise) vs. the CLI flow: the decision
- 7. CI checks and the portfolio artifact
- 8. Final project: version-controlled, documented delivery
Common questions
What people usually ask
No limit. It's a free guide: come in whenever you like, as often as you like.
No. Modules run from easier to harder, but you can jump to the one you need. Progress is saved per lesson.
Whatever is needed is listed under “What you need to bring”, above. If nothing is listed there, you can start from zero.
In the Club's WhatsApp group, and every two weeks there's a live with an instructor where questions get worked through.
Yes. It's issued automatically once you finish every lesson, with a verifiable code you can share on LinkedIn.
Start whenever you like
What students say
These reviews are from enrolled students who completed at least 50% of the course. We moderate reviews only on content grounds (spam, offensive language, personal data), never for being critical or negative.
No approved reviews yet.
Be the first to share your experience!