Google Antigravity agents get full context with GitLab Orbit

2026-06-25 · Source: GitLab · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Software Development & Engineering · Depth: Intermediate, short

Summary

Google Antigravity agents can now access GitLab Orbit, a lifecycle context graph, directly from the Antigravity MCP Store. This integration provides agents with structured access to projects, pipelines, merge requests, vulnerabilities, and source code across a GitLab instance. GitLab Orbit indexes the instance to build a knowledge graph of relationships, surfacing it through "query_graph" and "get_graph_schema" MCP tools. This enhanced context allows Antigravity agents to answer complex software lifecycle questions, such as module dependencies, unresolved vulnerabilities, and optimal merge request reviewers. Early internal tests showed 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. Key user journeys improved include blast radius analysis, codebase exploration for onboarding, and dynamic dependency mapping with image generation via Nano Banana Pro. Orbit reindexes within minutes and supports Ruby, Java, Kotlin, Python, TypeScript, JavaScript, Rust, and C# code.

Key takeaway

For MLOps Engineers or AI Engineers developing agents within Google Antigravity, integrating GitLab Orbit is crucial for enhancing agent intelligence and efficiency. This integration provides your agents with a live, structured understanding of your GitLab instance's software lifecycle, enabling them to perform complex queries and reduce hallucinations. You should install Orbit from the Antigravity MCP Store to leverage this context for tasks like blast radius analysis, dynamic dependency mapping, and improved developer onboarding, especially if you are on GitLab Premium or Ultimate tiers.

Key insights

Integrating a comprehensive software lifecycle knowledge graph significantly enhances AI agent accuracy and efficiency.

Principles

• Contextual data improves agent performance metrics.
• Real-time indexing prevents stale knowledge.
• Standardized protocols enable tool interoperability.

Method

GitLab Orbit indexes a GitLab instance to create a knowledge graph, which Antigravity agents query using JSON DSL via "query_graph" and "get_graph_schema" MCP tools.

In practice

• Perform blast radius analysis before code refactoring.
• Generate dynamic architecture diagrams from live data.
• Expedite developer onboarding with current codebase maps.

Topics

• Google Antigravity
• GitLab Orbit
• Knowledge Graphs
• AI Agents
• DevSecOps
• Software Lifecycle Management
• Model Context Protocol

Best for: AI Architect, CTO, VP of Engineering/Data, AI Engineer, Software Engineer, MLOps Engineer

Related on AIssential

• Google’s Antigravity Signals a Shift Beyond the IDE — Software development is shifting from human-centric coding in IDEs to agent-driven orchestration and supervision.
• From Ticket to First Commit: Building a Graph-Powered Contribution Strategist with GitLab Orbit — Knowledge graphs excel at contextualizing software development relationships beyond simple search.
• Agentic coding is only as good as its context — Agentic coding's effectiveness hinges on robust lifecycle context, enhancing speed and review processes.
• Introducing GitLab Orbit: Full code and lifecycle context, in one query — GitLab Orbit accelerates software agents and expands their query capabilities by providing comprehensive code and lifecycle context.
• Developing watsonx Orchestrate Agents with Antigravity — Google Antigravity can build IBM watsonx Orchestrate agents using open standards like MCP.
• Perplexity with Google Antigravity - Perplexity — Integrate Perplexity's Agent API into Google Antigravity projects for web-aware, grounded answers using an OpenAI-compatible interface.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by GitLab.