GitLab Orbit Now Powers Google Antigravity AI Agents
GitLab Orbit has landed inside Google Antigravity. Developers can now install GitLab’s context engine straight from the Antigravity MCP Store. As a result, coding agents finally get real visibility into a developer’s full software lifecycle.
So, what changed? Previously, Antigravity agents could only see open files and a terminal. They had no idea which services depended on the code being edited. They also could not tell if similar bugs had appeared before. Developers had to bridge that gap manually, often through copy-pasting between tools. That context lives in your DevSecOps platform, and getting it to a coding agent has meant using custom scripts or copy-pasting between tools.
GitLab Orbit changes that. It builds a live knowledge graph of projects, merge requests, pipelines, vulnerabilities, and source code. Then, it surfaces all of this through two MCP tools. One runs structured queries. The other returns the graph’s schema. Consequently, agents can now answer tougher questions instantly. For instance, they can flag which projects will break from a change. They can also surface unresolved vulnerabilities or suggest the right reviewer for a merge request.
The results are impressive. In early internal tests, agents grounded with GitLab Orbit responded up to 11 times faster, used up to 4.5 times fewer tokens, and produced up to 45 times fewer hallucinations. That is a massive leap in both speed and accuracy.
For Nigerian dev teams, this matters a lot. Many local engineering teams run lean. Often, one or two senior developers manage sprawling codebases inherited from contractors or earlier hires. GitLab Orbit’s onboarding feature could save serious time here. Instead of digging through outdated wikis, a new hire can simply ask Antigravity what a service depends on. The agent will respond with a live, accurate map. This is especially useful for Lagos-based fintech and e-commerce startups, where fast onboarding directly affects product velocity.
Installing Orbit is simple too. Open the MCP Store from Antigravity’s settings. Next, find the MCP section under customization. Then, click “Add MCP,” select GitLab Orbit, and authenticate. No terminal setup is required. Afterward, Orbit’s tools become available to your agents immediately.
Notably, this is the same engine behind GitLab’s Duo Agent Platform. Therefore, platform teams managing large GitLab instances do not need a separate context pipeline. Orbit indexes code across Ruby, Java, Kotlin, Python, TypeScript, JavaScript, Rust, and C#. Moreover, it reindexes within minutes of any change, keeping everything current.
GitLab Orbit is available on Premium and Ultimate tiers on GitLab.com. To get started, turn on Orbit for your top-level group, then install it from the Antigravity MCP Store. Teams new to this can also explore GitLab Duo Agent Platform’s free trial.
As AI agents take on more real engineering work, context will separate the useful tools from the noisy ones. For Nigerian teams scaling their AI-assisted workflows, GitLab Orbit is worth a serious look.
