Agents at work: plan - test - CR - deploy - repeat
summary
GitLab integrates Amazon Q Developer agents into the repo and MR UI so agents run via review agent, dev agent, and q transform using slash commands; agents break a huge backlog item into substories, estimate it (Q returns 13 story points), and create sub-tasks. Review agent flags static-code weaknesses, software composition vulnerabilities, and risky changes (breaking APIs, subprocess/eval usage) and can auto-apply simple fixes; everything ties into CI pipelines so tests run after agent edits. Preview is available for self‑hosted GitLab Ultimate with a 60‑day trial.
AI-written summary · about
chapters (6)
- 0:00Introduction to AWS Dev Day
- 2:05Integrating Q in GitLab
- 3:42Backlog management with Q
- 8:48Automated code review
- 12:20Bug fixing and automation
- 18:06Getting started and closing
related talks
31mAgents at Work: How GitLab Is Redefining AI Adoption for Engineering Teams
36mAgentic CI/CD: from Pull Request to Production without the Paper Cuts
40mHow To Tame Your Agents?
22mThe place of GitLab within a mobile team's processes
24mReinvent your pipeline with GitLab, Kubernetes, and Amazon EKS
24mEmpowering the GitOps Future with AI
27mIntegrating Agentic AI Into Internal Developer Platforms To Minimize Errors and Downtime
30mBuilding Tools with Tools: Modern AI-Driven Development