tech
GitLab 19.0 targets the gap between writing code and shipping it
GitLab 19.0 extends agentic AI across the full software lifecycle with its Duo Agent Platform, adds SBOM-based dependency scanning, and supports Claude Opus 4.7 and Gemini models. The release targets the gap between faster code generation and slower delivery pipelines.
TL;DR
- GitLab 19.0 focuses on 'intelligent orchestration' to automate manual processes in the software lifecycle.
- The Duo Agent Platform now spans the entire software lifecycle, running tasks in parallel.
- A new SBOM-based dependency scanner provides visibility into vulnerabilities in declared and transitive dependencies.
- GitLab Duo Developer offers more flexible AI trigger methods, allowing agents to pick up work autonomously.
- The release adds support for Claude Opus 4.7, Google's Gemini models, and open-source options for self-hosted deployments.
- Group-level custom review instructions reduce setup overhead across multiple repositories.
- Infrastructure changes include replacing Redis with Valkey and dropping Ubuntu 20.04 support.
- GitLab Credits are introduced to meter AI agent usage, with different allowances for Premium and Ultimate customers.
- The company restructured R&D into autonomous teams to support the AI strategy.
- GitLab aims to have a structural advantage with a single platform covering planning to deployment, rather than point solutions.