GitLab’s latest release uses AI for more than just making code completion smarter, with customisable agents, a Knowledge Graph, and more.
A developer’s day is often a constant juggle. You’re not just writing code; you’re deciphering legacy systems, navigating complex dependencies, wrestling with pipelines, and trying to remember your team’s specific, unwritten rules for formatting bug reports. The promise of AI has been to help, but often it feels like just another tool to manage.
GitLab’s latest release, version 18.4, attempts to change that daily juggle by fundamentally rethinking our relationship with AI. The big idea? To stop treating AI like a parrot on your shoulder and start treating it like a custom-built member of your team.
Build a custom AI agent to be the exact colleague you need
The real headline feature is GitLab’s new AI Catalog. Think of it less as a store and more as a training ground for your own digital apprentices. Every team has its own unique way of doing things, and now you can bake that institutional knowledge directly into an AI.
Imagine you could spin up a ‘Security Specialist’ agent that knows your company’s compliance rules inside and out and automatically flags issues in every merge request. Or what about a ‘Documentation Pro’ agent that takes your rough notes and instantly drafts perfectly formatted guides in your team’s preferred style? You could even create a ‘Bug Squasher’ that files reports with the precise labels and format your project manager insists on, every single time.
This is about taking the repetitive tasks that drain your focus and delegating them to an AI that does them your way. They slot right into your existing workflow, becoming a natural part of how you get things done rather than another distracting app.
GitLab Knowledge Graph gives your AI a map of your code
Of course, an AI teammate is only useful if it actually understands the project. We’ve all had a chatbot confidently give us nonsense because it lacks context. That’s where the new GitLab Knowledge Graph comes in.
The new GitLab Knowledge Graph creates a living, breathing map of your entire codebase. It doesn’t just see files; it understands the connections between them like how a change in one area might ripple out and affect another. You can ask it things like, “What other services does this change impact?” and get a real answer. For your new GitLab AI agents, it’s the difference between wandering in the dark and having a complete architectural blueprint. It gives them the deep context they need to provide helpful insights.
A safety net that understands your priorities
Even with the best teammates, things still break. We’ve all seen that dreaded red ‘X’ on a pipeline and felt that sinking feeling.
The new ‘Fix Failed Pipelines Flow’ is designed to be business-aware—understanding that a broken pipeline on the main customer-facing app is a five-alarm fire, while a failure in an experimental feature branch can probably wait. It analyses the error logs with this context in mind, identifies the root cause, and even suggests a fix by creating a merge request. It’s like having a senior engineer who not only spots the problem but also understands its urgency and helps you sort it.
Okay, this all sounds great, but letting AI loose on your proprietary code can feel risky. GitLab seems to have anticipated this anxiety by building in much-needed guardrails. GitLab provides granular control to select which large language models power your AI features, ensuring you can stick with providers you trust.
Even better is the new Context Exclusion feature. This lets you draw red lines around sensitive parts of your project – think files with secrets, passwords, or core proprietary algorithms – and tell the AI, “You cannot access this. Ever.” It gives you the confidence to take advantage of AI’s power without compromising your security.
GitLab’s latest release promises a future where you’re not just a coder working on a single problem, but an orchestrator collaborating with a team of specialised AI agents – backed with the context of the new Knowledge Graph – to solve problems faster and more effectively.
See also: XZ attack reveals unlearned open-source security lessons
Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.
AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.