GitLab
This section covers GitLab as a delivery platform: working on a shared repository as a contributor, building and deploying with GitLab CI/CD, governing that repository as a maintainer, running the instance as an administrator, scaling merges with merge trains, and using GitLab Duo’s AI features in day-to-day work.
The pages are grouped by role. Read the path below in order the first time; afterwards, jump to the stage you need.
Learning path
Section titled “Learning path”Stage 1: Contributor basics
Section titled “Stage 1: Contributor basics”The everyday workflow for anyone pushing changes to a GitLab project. Everything later assumes you can do this.
- Contributor basics — clone over SSH or HTTPS, create merge requests (including push options), and resolve merge conflicts in the UI or on the command line.
Stage 2: CI/CD pipelines
Section titled “Stage 2: CI/CD pipelines”Learn the concepts first, keep the keyword reference open, then build and deploy.
- GitLab CI/CD Pipelines overview — pipelines, stages, jobs, runners, variables, artifacts and caching, with a complete
.gitlab-ci.ymlexample. It lives in the DevOps section and is the starting point for this stage. - GitLab CI/CD Keywords and Usage — a reference for stages, jobs and
rules/whenconditions such asalways,neverandon_success. Skim it once, then come back to it while you write pipelines. - GitLab Build Pipeline — building front-end, .NET service and database artifacts.
- GitLab Deploy Pipeline — promoting those artifacts across environments.
Every job runs on a GitLab Runner and reads its settings and secrets from CI/CD variables. Administrators usually set both up (stage 4), but read that section now if your team owns its own runners or pipeline secrets. For schedules, notifications, failure handling and tool integrations, see Pipeline guides in the DevOps section below.
Stage 3: Maintainer tasks
Section titled “Stage 3: Maintainer tasks”Protect the main branch and make reviews and history trustworthy before many people contribute. The GitLab administration guides list stages 3 and 4 on one page.
- Protected branches, approvals and code owners — branch rules and wildcards, push and merge permissions, approval rules,
CODEOWNERSsections, and merge request and issue templates. - Push rules and server hooks — branch naming rules, commit message formats, required Jira keys, secret and file rules, applied to every push.
Stage 4: Administrator tasks
Section titled “Stage 4: Administrator tasks”Instance- and group-level setup that CI/CD and governance depend on.
- Runners, CI/CD variables and webhooks — registering runners with authentication tokens, masked and protected variables, and signed webhooks.
- Pre-receive server hooks — custom server-side checks on GitLab Self-Managed when push rules are not enough.
Stage 5: Scaling delivery with merge trains
Section titled “Stage 5: Scaling delivery with merge trains”- GitLab Merge Train — how merge trains differ from a standard merge, and when a busy main branch needs them.
Stage 6: AI assistance with GitLab Duo
Section titled “Stage 6: AI assistance with GitLab Duo”- GitLab Duo overview — the Agent Platform, non-agentic features such as Code Review and Root Cause Analysis, add-ons and credits.
- GitLab Duo Agent — agentic workflows inside GitLab.
Pipeline guides in the DevOps section
Section titled “Pipeline guides in the DevOps section”These pages cover GitLab CI/CD techniques and integrations in more depth. Read them after stage 2.
| Guide | What it adds |
|---|---|
| GitLab Pipelines introduction | Topics at a glance: links to the GitLab pipeline guides |
| Pipeline Templates | Reusing jobs with YAML anchors, extends and include |
| Failure Handling | Retries, allow_failure and rules |
| Scheduled Pipelines | Running pipelines on cron schedules |
| Notifications | Email and Slack notifications configured in YAML |
| Ansible from GitLab CI | Running Ansible playbooks from a pipeline |
| Liquibase in GitLab CI | Validating and applying database changes in CI |
| TeamCity and GitLab | Webhooks, status reporting and merge request checks when TeamCity runs the builds |
Where to go next
Section titled “Where to go next”- GitLab CI/CD prompts — ready-to-use prompts for explaining, debugging and documenting pipelines.
- Kiro + GitLab MCP — give an AI editor read access to issues, merge requests and pipelines.
- GitLab certifications — exam paths if you want to formalize the skills above.
- Git fundamentals, branching strategies and merge vs rebase — the Git concepts every GitLab workflow relies on.
- GitHub — the equivalent guides for GitHub.