GitLab Duo Agent Platform is a new AI-powered solution that embeds multiple intelligent assistants ("agents") throughout your software development lifecycle. It serves as an orchestration layer where developers collaborate asynchronously with AI agents across DevSecOps, transforming linear workflows into dynamic, parallel processes.
Routine tasks, from code refactoring and security scans to research, can be delegated to specialized AI agents, freeing human developers to focus on solving complex problems and driving innovation.
The platform leverages GitLab's role as a central DevSecOps platform (encompassing code management, CI/CD pipelines, issue tracking, test results, security scans, and more) to provide these agents with complete project context, enabling them to contribute meaningfully while adhering to your team's standards and practices.
This comprehensive eight-part guide will take you from your first interaction to production-ready automation workflows with full customization.
💡 Join GitLab Transcend on February 10 to learn how agentic AI transforms software delivery. Hear from customers and discover how to jumpstart your own modernization journey. Register now.
Evolution from GitLab Duo Pro/Enterprise to Duo Agent Platform
GitLab Duo Agent Platform is an evolution, not a replacement of Duo Pro and Enterprise. It's a superset that moves from 1:1 developer-AI interactions to many-to-many team-agent collaboration.
- Duo Pro enhanced individual developer productivity in the IDE with AI-powered code suggestions and chat.
- Duo Enterprise expanded beyond coding to deliver comprehensive AI capabilities across the entire software development lifecycle. But it was still primarily an approach in enabling 1:1 interaction between the user and an AI assistant — mostly a Q&A experience with one use case at a time.
- Duo Agent Platform moves from 1:1 interactions to many-to-many team-agent collaboration, where specialized agents autonomously handle routine tasks across the software lifecycle.
The complete series
| Part | Title | What You'll Learn |
|---|---|---|
| 1 | Introduction to GitLab Duo Agent Platform | Platform architecture, four ways to use agents, accessing agents and flows, first interactions, sessions, and model selection |
| 2 | Getting started with GitLab Duo Agentic Chat | Accessing chat across Web UI and IDEs, model selection and switching, agent selection, common use cases, and troubleshooting |
| 3 | Understanding agents: Foundational, custom, and external | Foundational agents (GitLab Duo, Planner, Security Analyst, Data Analyst), creating custom agents with system prompts, external agents setup, AGENTS.md customization, and choosing the right agent type |
| 4 | Understanding flows: Multi-agent workflows | Introduction to foundational flows, creating custom YAML workflows, flow execution, multi-agent orchestration, and monitoring |
| 5 | AI Catalog: Discover, create, and share agents and flows | Browsing and discovering agents and flows, enabling agents and flows in projects, creating and publishing your own agents and flows, and managing visibility |
| 6 | Monitor, manage, and automate AI workflows | Automate menu overview, monitoring sessions with detailed logs, setting up event-driven triggers, and managing AI workflows |
| 7 | Model Context Protocol integration | MCP overview, GitLab as MCP client connecting to external tools, GitLab as MCP server for external AI tools, and configuration examples |
| 8 | Customizing GitLab Duo Agent Platform | Custom chat rules, AGENTS.md configuration, system prompts for agents, agent tool configuration, MCP setup, and custom flow YAML configuration |
Key concepts reference
Core components
| Component | Description | Key Features |
|---|---|---|
| Duo Agentic Chat | Primary interface for agent interaction | • Available in Web UI and IDEs • Supports model selection • Maintains conversation history |
| Agents | Specialized AI collaboration partners for specific tasks | • Foundational: Provided by GitLab (Planner, Security Analyst, etc.) • Custom: Created by your team • External: External AI providers like Claude and OpenAI |
| Flows | Multi-step workflows combining agents | • Foundational: Provided by GitLab (Developer, Fix CI/CD Pipeline, etc.) • Custom: User-defined workflows you create |
| AI Catalog | Central repository for discovering, creating, and sharing | • Browse and discover agents and flows • Add to your projects • Share across organization |
| Automate Menu | Management hub for AI workflows | • Sessions: Flow activity logs • Flows: Multi-step workflows • Agents: Specialized AI assistants • Triggers: Event-based automation |
| Model Context Protocol (MCP) | External integration framework | • Client: GitLab Duo connects to external MCP servers (Jira, Slack, AWS, etc.) • Server: GitLab acts as MCP server for external AI tools (Claude Desktop, Cursor, etc.) |
Essential terminology
| Term | Definition |
|---|---|
| Agent | Specialized AI assistant for specific tasks and to answer complex questions |
| Foundational Agent | Pre-built agents created and maintained by GitLab (e.g., GitLab Duo, Planner, Security Analyst) — available immediately with no setup |
| Custom Agent | Agents you create with custom system prompts and tools for team-specific workflows — configured through project/group settings |
| External Agent | External AI providers like Claude, OpenAI, Google Gemini, and more integrated into the platform |
| Flow | Combination of one or more agents working together to solve a complex problem |
| Foundational Flow | Pre-built workflows by GitLab (Issue to MR, Fix Pipeline, Convert Jenkins, Software Development Flow) — triggered via UI buttons or IDEs |
| Custom Flow | YAML-defined workflows you create for team-specific automation — triggered by events or mentions |
| Trigger | Event that automatically starts a flow (e.g., mention, assignment) |
| Session | Record of agent or flow activity with complete logs and pipeline execution details |
| System Prompt | Instructions defining agent behavior, expertise, and communication style |
| Service Account | Account used by flows or external agents to perform GitLab operations with specific permissions |
| MCP | Model Context Protocol for external integrations (connects to Jira, Slack, AWS, etc.) |
| AGENTS.md | Industry-standard file for customizing agent behavior at user or workspace level |
| Custom Rules | Rules that customize how GitLab Duo behaves in your IDE |
| Tools | Capabilities that agents can use to interact with GitLab and external systems (e.g., create issues, merge requests, run pipelines, analyze code) |
Ready to get started?
Begin your journey with Part 1: Introduction to GitLab Duo Agent Platform to learn the platform fundamentals.
Feedback
We'd love to hear from you! Found an error? Have a suggestion?