GitLab 19.0: Bridging the Gap Between Code and Shipment
GitLab has released GitLab 19.0, its first major update in a year, focusing on intelligent orchestration to revolutionize software delivery. The platform aims to address the discrepancy between faster code generation assisted by AI and the slower manual processes that follow.
Key Features:
- Agentic AI across the Lifecycle: GitLab 19.0 expands the Duo Agent Platform, enabling AI to assist in every stage of development, from planning to security remediation.
- SBOM-based Dependency Scanning: The new capability offers full visibility into vulnerabilities within dependency trees, addressing a significant portion of critical security debt (70% according to Veracode).
- Enhanced Duo Developer: Developers can now trigger the AI assistant in various ways, including assigning it to issues or mentioning it in discussions, promoting autonomous work.
- Support for Advanced Models: The update includes support for Claude Opus 4.7, Google’s Gemini models, open-source alternatives like Devstral 2 and GLM-5.1, and Mistral AI.
- Group-Level Custom Review Instructions: Simplifying setup, a single set of instructions can now apply across an entire group and its subgroups.
- Infrastructure Updates: Introducing Valkey as the new default in the Linux package and removing bundled Mattermost are among the changes, but some require planning for self-managed customers upgrading from version 18.
GitLab’s approach aims to streamline software delivery by automating manual bottlenecks, ensuring faster and more efficient code shipment.