Agentic AI Cookbook · GitLab Well Architected

Guides to custom and foundational flows, with their examples, included skills, and evaluation material kept together. Each guide walks from the scenario through setup and operation without depending on another catalog entry.
Find CI configurations that need attention, review proposed migrations, and plan the changes with the teams who own those pipelines.
Review a merge request against your project's compliance policy, with an explanation for each result, any checks still needed, and a suggested label.
Investigate a code change, review the plan, and use each implementation's code and tests to direct the next increment.
Find guides and runbooks left out of date by a code change, approve the suggested edits, and review the checked updates in a separate merge request.
Follow a failed pipeline through Duo's diagnosis, review its suggested fix, and check the next pipeline. Learn when a project setting is the better answer.
Compare a proposed Flow change with a version you know works, review the test results, and decide whether to adopt it or restore the earlier version.
Find projects in a GitLab group whose CI includes don't match your team's policy, then review the findings before creating issues for their owners.
Use incident observations and GitLab records to understand what happened and plan work to prevent it from happening again, while leaving production operations to your team.
Check whether an issue gives an agent enough information to start work, with suggestions for missing details grounded in the repository.
Organize a migration across projects, approve each group of projects before its work begins, and track progress through their issues.
Compare Duo's review before and after adding project-specific instructions, then fix the issue and check the next review.
Turn a clear issue into a tested merge request, with a person reviewing the plan before implementation begins.
Review an advisory across current dependency scans, keep missing checks visible, and approve follow-up issues before project owners make and verify repairs.
Use GitLab Duo with Orbit to find callers, check what a library change means for them, and choose a useful regression test.
Delegate routine implementation to AI. Get working code with tests from requirements.
Get consistent, thorough code reviews at scale. Enforce patterns across all MRs automatically.
Enforce team code style consistently across all MRs. Eliminate bikeshedding in reviews.
Review database migrations for safety, reversibility, and performance impact. Prevent production issues.
Ensure API documentation is complete with examples, error responses, and authentication details.
Implement AWS-native features with best practices, infrastructure code, and compliance requirements.
Identify related open MRs that might conflict. Coordinate changes across team members.
Catch compliance violations early before they block deployment. Pre-check for HIPAA, SOC2, GDPR.
Analyze test coverage gaps and suggest specific test cases. Ensure critical paths are tested.
Think through approaches collaboratively. Get challenged on tradeoffs and edge cases you might miss.
Document how your system integrates with internal and external APIs. Prevents breaking changes in integrations.
Understand complex database schemas, relationships, and optimization opportunities. Identify performance bottlenecks in your data model.
Triage security scan results by real risk vs false positives. Prioritize remediation effectively.
Parse complex stack traces to identify error origins and suggest fixes.
Challenge architectural assumptions collaboratively. Explore failure modes and scaling concerns.
Evaluate if an MR includes adequate tests for new functionality. Identify gaps before they reach production.
Assess external dependencies for vulnerabilities, outdated versions, and license compliance.
Audit codebases for compliance violations. Check data handling, access controls, and encryption.
Automate repetitive development tasks with error handling and logging. Eliminate manual toil.
Transform large epics into manageable, well-scoped issues with clear dependencies and acceptance criteria.
Scaffold complete applications with database schemas, API endpoints, and frontend components.
Categorize features based on MoSCoW criteria.
Ensure code follows team conventions and style guides. Consistent enforcement without tedious manual checks.
Understand design decisions and tradeoffs in MRs. Get context beyond just what changed.
See which programming languages have the highest code suggestion acceptance rates to understand where AI assistance is most effective.