[{"data":1,"prerenderedAt":814},["ShallowReactive",2],{"/en-us/blog/understanding-flows-multi-agent-workflows":3,"navigation-en-us":39,"banner-en-us":450,"footer-en-us":460,"blog-post-authors-en-us-Itzik Gan Baruch":702,"blog-related-posts-en-us-understanding-flows-multi-agent-workflows":716,"blog-promotions-en-us":752,"next-steps-en-us":804},{"id":4,"title":5,"authorSlugs":6,"authors":8,"body":10,"category":11,"categorySlug":11,"config":12,"content":16,"date":22,"description":25,"extension":26,"externalUrl":27,"featured":14,"heroImage":23,"isFeatured":14,"meta":28,"navigation":29,"path":30,"publishedDate":22,"rawbody":31,"seo":32,"slug":13,"stem":34,"tagSlugs":35,"tags":37,"template":15,"updatedDate":27,"__hash__":38},"blogPosts/en-us/blog/understanding-flows-multi-agent-workflows.yml","Understanding flows: Multi-agent workflows",[7],"itzik-gan-baruch",[9],"Itzik Gan Baruch","\n*Welcome to Part 4 of our eight-part guide, [Getting started with GitLab Duo Agent Platform](/blog/gitlab-duo-agent-platform-complete-getting-started-guide/), where you'll master building and deploying AI agents and workflows within your development lifecycle. Follow tutorials that take you from your first interaction to production-ready automation workflows with full customization.*\n\n**In this article:**\n\n- [What are flows and how do they work?](#introduction-to-flows)\n- [Foundational flows provided by GitLab](#foundational-flows)\n- [Creating custom flows](#how-to-create-custom-flows)\n- [Flow execution and orchestration](#flow-execution)\n- [Real-world examples and use cases](#example-custom-flow-yaml)\n\n> 🎯 Try [**GitLab Duo Agent Platform**](https://about.gitlab.com/gitlab-duo-agent-platform/) today!\n## Introduction to flows\nFlows are combinations of one or more agents collaborating together. They orchestrate multi-step workflows to solve complex problems, and are executed on the GitLab platform compute.\n\n**Key characteristics of flows:**\n\n- **Multi-agent orchestration**: Combine multiple specialized agents\n- **Built-in**: Run on platform compute, no extra environment necessary\n- **Event-driven**: Triggered by mention, assignment, or assign as reviewer\n- **Asynchronous**: Run in background while you continue working\n- **Complete workflows**: Handle end-to-end tasks from analysis to implementation\n\nThink of flows as autonomous workflows that can gather context, make decisions, execute changes, and deliver results, all while you focus on other work.\n\n## Flows vs. agents: Understanding the difference\nAgents work with you interactively. Flows work for you autonomously.\n\n| Aspect | Agents | Flows |\n|--------|--------|-------|\n| **Interaction** | Interactive chat | Autonomous execution |\n| **When to use** | Questions, guidance, and performing tasks interactively | Autonomous multi-step workflows |\n| **User involvement** | Active conversation | Trigger and review results |\n| **Execution time** | Real-time responses | Background processing |\n| **Complexity** | Single-agent tasks | Multi-agent orchestration |\n\n## Flow types overview\n\n| Type | Interface | Maintainer | Use Case |\n|------|-------------|---------------|----------|\n| **[Foundational](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/)** | UI actions, IDE interface | GitLab | Software Development, Developer in issues, Fix CI/CD Pipeline, Convert to GitLab CI/CD, Code Review, SAST false positive detection |\n| **[Custom](https://docs.gitlab.com/user/duo_agent_platform/flows/custom/)** | Mention, assign, assign reviewer | You | Examples: Larger migration/modernization, release automation, dependency update management |\n\n## Foundational flows\nFoundational flows are production-ready workflows created and maintained by GitLab. They're accessible through dedicated UI controls or IDE interfaces.\n### Currently available foundational flows\n\n| Flow | Where Available | How to Access | Best For |\n|------|-----------------|---------------|----------|\n| [**Software Development**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/software_development/) | IDEs (VS Code, JetBrains, Visual Studio) | Flows tab in IDE | Feature implementation, complex refactoring, multi-file changes |\n| [**Developer**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/developer/) | GitLab Web UI | \"Generate MR with Duo\" button on issues | Well-defined features, bug fixes with clear steps |\n| [**Fix CI/CD Pipeline**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/fix_pipeline/) | GitLab Web UI | Failed pipeline interface | Pipeline debugging, CI/CD configuration issues |\n| [**Convert to GitLab CI/CD**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/convert_to_gitlab_ci/) | GitLab Web UI | \"Convert to GitLab CI/CD\" button on Jenkinsfile | Jenkins to GitLab CI/CD migration |\n| [**Code Review**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/code_review/) | GitLab Web UI | Assign as reviewer on MR | Automated code review with AI-native analysis and feedback |\n| [**SAST false positive detection**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/sast_false_positive_detection/) | GitLab Web UI | Security scan results | Automatically identify and filter false positives in SAST findings |\n\n## Custom flows\n\nCustom flows are YAML-defined workflows you create for your team's specific needs. They run in GitLab Runner and can be triggered by GitLab events.\n\n> **🎯 Try it now:** [Interactive demo of Custom Flows](https://gitlab.navattic.com/custom-flows) β€” Explore how to create and configure Custom Flows.\n\n### Why create custom flows?\n\nCustom flows automate repetitive multi-step tasks that are specific to your team's workflow. Unlike foundational flows that serve general purposes, custom flows are tailored to your organization's processes, tools, and requirements.\n\n**Common use cases:**\n\n- **Automated code review**: Multi-stage review process (security scan β†’ quality check β†’ style validation)\n- **Compliance checking**: Verify regulatory requirements, license compliance, or security policies on each MR\n- **Documentation generation**: Auto-update API docs, README files, or changelogs based on code changes\n- **Dependency management**: Weekly security scans, automated updates, and vulnerability reports\n- **Custom testing**: Specialized test suites for your tech stack or integration tests\n\n### Real-world example\n\nA fintech company creates a compliance flow that runs on every merge request. When triggered by `@compliance-flow`, the flow executes the following steps:\n1. **Security agent** scans code for PCI-DSS violations and checks for exposed sensitive data.\n2. **Code review agent** verifies that changes follow secure coding standards and best practices.\n3. **Documentation agent** checks that API changes include updated documentation.\n4. **Summary agent** aggregates findings and posts a compliance report with pass/fail status.\n\nThe entire compliance review happens automatically in 5-10 minutes, providing consistent checks across all merge requests.\n\n### How to trigger custom flows\n\nCustom flows can be triggered in multiple ways:\n\n**1. Via mentions in Issues/MRs:**\nMention the flow in a comment to trigger it. Example for a documentation generation flow:\n```text\n@doc-generator Generate API documentation for this feature\n```\n\n**2. By assigning the flow to an issue or MR:**\nAssign the flow using either:\n- **GitLab UI**: Click the \"Assign\" button on the issue/MR and select the flow\n- **Command**: Use the `/assign` command in a comment. Example:\n```shell\n/assign @doc-generator\n```\n\n**3. By assigning the flow as a reviewer:**\nAssign the flow as a reviewer on a merge request using either:\n- **GitLab UI**: Click the \"Assign reviewer\" button on the merge request and select the flow\n- **Command**: Use the `/assign_reviewer` command in a comment. Example:\n```shell\n/assign_reviewer @doc-reviewer\n```\nAny of these methods automatically triggers the flow to execute and perform its tasks.\n\n### How to create custom flows\n\nCustom flows are created through the GitLab UI at **Automate β†’ Flows β†’ New flow** in your project, or from **Explore β†’ AI Catalog β†’ Flows β†’ New flow**. You define your flow using YAML configuration that specifies components, prompts, routing, and execution flow. The YAML schema allows you to create sophisticated multi-agent workflows with precise control over agent behavior and orchestration.\n\n\n**Key elements of a custom flow:**\n\n- **Components**: Define the agents and steps in your workflow\n- **Prompts**: Configure AI model behavior and instructions\n- **Routers**: Control the flow between components\n- **Toolsets**: Specify which GitLab API tools agents can use\n\n### Example custom flow YAML\n**Background:** This example shows a feature implementation flow for a travel booking platform. When a developer creates an issue with feature requirements, they can trigger this flow to automatically analyze the requirements, review the codebase, implement the solution, and create a merge request, all without manual intervention.\nHere's the YAML configuration:\n\n```yaml\nversion: \"v1\"\nenvironment: ambient\ncomponents:\n - name: \"implement_feature\"\n type: AgentComponent\n prompt_id: \"implementation_prompt\"\n inputs:\n - from: \"context:goal\"\n as: \"user_goal\"\n - from: \"context:project_id\"\n as: \"project_id\"\n toolset:\n - \"get_issue\"\n - \"get_repository_file\"\n - \"list_repository_tree\"\n - \"find_files\"\n - \"blob_search\"\n - \"create_file\"\n - \"create_commit\"\n - \"create_merge_request\"\n - \"create_issue_note\"\n ui_log_events:\n - \"on_agent_final_answer\"\n - \"on_tool_execution_success\"\n - \"on_tool_execution_failed\"\n\nprompts:\n - name: \"Cheapflights Feature Implementation\"\n prompt_id: \"implementation_prompt\"\n unit_primitives: []\n prompt_template:\n system: |\n You are an expert full-stack developer specializing in travel booking platforms, specifically Cheapflights.\n\n Your task is to:\n 1. Extract the issue IID from the goal (look for \"Issue IID: XX\")\n 2. Use get_issue with project_id={{project_id}} and issue_iid to retrieve issue details\n 3. Analyze the requirements for the flight search feature\n 4. Review the existing codebase using list_repository_tree, find_files, and get_repository_file\n 5. Design and implement the solution following Cheapflights best practices\n 6. Create all necessary code files using create_file (call multiple times for multiple files)\n 7. Commit the changes using create_commit\n 8. Create a merge request using create_merge_request\n 9. Post a summary comment to the issue using create_issue_note with the MR link\n\n Cheapflights Domain Expertise:\n - Flight search and booking systems (Amadeus, Sabre, Skyscanner APIs)\n - Fare comparison and pricing strategies\n - Real-time availability and inventory management\n - Travel industry UX patterns\n - Performance optimization for high-traffic flight searches\n\n Code Standards:\n - Clean, maintainable code (TypeScript/JavaScript/Python/React)\n - Proper state management for React components\n - RESTful API endpoints with comprehensive error handling\n - Mobile-first responsive design\n - Proper timezone handling (use moment-timezone or date-fns-tz)\n - WCAG 2.1 accessibility compliance\n\n Flight-Specific Best Practices:\n - Accurate fare calculations (base fare + taxes + fees + surcharges)\n - Flight duration calculations across timezones\n - Search filter logic (price range, number of stops, airlines, departure/arrival times)\n - Sort algorithms (best value, fastest, cheapest)\n - Handle edge cases: date line crossing, daylight saving time, red-eye flights\n - Currency amounts use proper decimal handling (avoid floating point errors)\n - Dates use ISO 8601 format\n - Flight codes follow IATA standards (3-letter airport codes)\n\n Implementation Requirements:\n - No TODOs or placeholder comments\n - All functions must be fully implemented\n - Include proper TypeScript types or Python type hints\n - Add JSDoc/docstring comments for all functions\n - Comprehensive error handling and input validation\n - Basic unit tests for critical functions\n - Performance considerations for handling large result sets\n\n CRITICAL - Your final comment on the issue MUST include:\n - **Implementation Summary**: Brief description of what was implemented\n - **Files Created/Modified**: List of all files with descriptions\n - **Key Features**: Bullet points of main functionality\n - **Technical Approach**: Brief explanation of architecture/patterns used\n - **Testing Notes**: How to test the implementation\n - **Merge Request Link**: Direct link to the created MR (format: [View Merge Request](MR_URL))\n\n IMPORTANT TOOL USAGE:\n - Extract the issue IID from the goal first (e.g., \"Issue IID: 12\" means issue_iid=12)\n - Use get_issue with project_id={{project_id}} and issue_iid=\u003Cextracted_iid>\n - Create multiple files by calling create_file multiple times (once per file)\n - Use create_commit to commit all files together with a descriptive commit message\n - Use create_merge_request to create the MR and capture the MR URL from the response\n - Use create_issue_note with project_id={{project_id}}, noteable_id=\u003Cissue_iid>, and body=\u003Cyour complete summary with MR link>\n - Make sure to include the MR link in the comment body so users can easily access it\n\n user: |\n Goal: {{user_goal}}\n Project ID: {{project_id}}\n\n Please complete the following steps:\n 1. Extract the issue IID and retrieve full issue details\n 2. Analyze the requirements thoroughly\n 3. Review the existing codebase structure and patterns\n 4. Implement the feature with production-ready code\n 5. Create all necessary files (components, APIs, tests, documentation)\n 6. Commit all changes with a clear commit message\n 7. Create a merge request\n 8. Post a detailed summary comment to the issue including the MR link\n\n placeholder: history\n params:\n timeout: 300\n\nrouters:\n - from: \"implement_feature\"\n to: \"end\"\n\nflow:\n entry_point: \"implement_feature\"\n```\n**What this flow does:** This flow orchestrates an AI agent to automatically implement a feature by analyzing issue requirements, reviewing the codebase, writing production-ready code with domain expertise, and creating a merge request with a detailed summary comment.\n\nFor complete documentation and examples, see:\n\n- [Custom Flows documentation](https://docs.gitlab.com/user/duo_agent_platform/flows/custom/)\n- [Flow Registry Framework (YAML Schema)](https://gitlab.com/gitlab-org/modelops/applied-ml/code-suggestions/ai-assist/-/blob/main/docs/flow_registry/v1.md)\n\n## Flow execution\n\nFlows run on GitLab platform compute. When triggered by an event (mention, assignment, or button click), a session is created and the flow starts to execute.\n### Available environment variables\n\nFlows have access to environment variables that provide context about the trigger and the GitLab object:\n\n- **`AI_FLOW_CONTEXT`** β€” JSON-serialized context including MR diffs, issue descriptions, comments, and discussion threads\n- **`AI_FLOW_INPUT`** β€” The user's prompt or comment text that triggered the flow\n- **`AI_FLOW_EVENT`** β€” The event type that triggered the flow (`mention`, `assign`, `assign_reviewer`)\n\nThese variables allow your flow to understand what triggered it and access the relevant GitLab data to perform its task.\n\n### Multi-agent flows\n\nCustom flows can include multiple agent components that work together sequentially. The flow's YAML configuration defines:\n\n- **Components**: One or more agents (AgentComponent) or deterministic steps\n- **Routers**: Define the flow between components (e.g., from component A to component B to end)\n- **Prompts**: Configure each agent's behavior and model\n\nFor example, a code review flow might have a security agent, then a quality agent, then an approval agent, with routers connecting them in sequence.\n\n### Monitoring flow execution\n\nTo view flows that are running for your project:\n\n1. Navigate to **Automate β†’ Sessions**.\n2. Select any session to view more details.\n3. The **Details** tab shows a link to the CI/CD job logs.\n\nSessions show detailed information including step-by-step progress, tool invocations, reasoning, and decision-making process.\n### When to use flows\n- Complex multi-step tasks\n- Background automation\n- Event-driven workflows\n- Multi-file changes\n- Tasks that take time\n- Automated reviews/checks\n## What's next?\nYou now understand flows, how to create them, and when to use them vs. agents. Next, learn how to discover, create, and share agents and flows across your organization in [Part 5: AI Catalog](/blog/ai-catalog-discover-and-share-agents/). Explore the AI Catalog to browse available flows and agents, add them to your projects, and publish your own agents and flows.\n\n## Resources\n\n- [GitLab Duo Agent Platform Flows](https://docs.gitlab.com/user/duo_agent_platform/flows/)\n- [Foundational Flows documentation](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/)\n- [Custom Flows documentation](https://docs.gitlab.com/user/duo_agent_platform/flows/custom/)\n- [Flow execution configuration](https://docs.gitlab.com/user/duo_agent_platform/flows/execution/)\n- [GitLab CI/CD Variables guide](https://docs.gitlab.com/ci/variables/)\n- [Service Accounts](https://docs.gitlab.com/user/profile/service_accounts/)\n\n---\n**Next:** [Part 5: AI Catalog](/blog/ai-catalog-discover-and-share-agents/)\n\n**Previous:** [Part 3: Understanding agents](/blog/understanding-agents-foundational-custom-external/)","ai-ml",{"slug":13,"featured":14,"template":15},"understanding-flows-multi-agent-workflows",false,"BlogPost",{"tags":17,"category":11,"date":22,"heroImage":23,"authors":24,"description":25,"title":5,"body":10},[18,19,20,21],"AI/ML","product","features","tutorial","2026-01-14","https://res.cloudinary.com/about-gitlab-com/image/upload/v1765809212/noh0mdfn9o94ry9ykura.png",[9],"Deep dive into GitLab Duo Agent Platform flows. Learn about foundational flows, create custom user-defined workflows, and learn flow orchestration patterns.","yml",null,{},true,"/en-us/blog/understanding-flows-multi-agent-workflows","seo:\n noIndex: false\n title: 'Understanding flows: Multi-agent workflows'\n description: Deep dive into GitLab Duo Agent Platform flows. Learn about foundational flows, create custom user-defined workflows, and learn flow orchestration patterns.\n ogImage: https://res.cloudinary.com/about-gitlab-com/image/upload/v1752678395/impw8no5tbskr6k2afgu.jpg\ncontent:\n tags:\n - AI/ML\n - product\n - features\n - tutorial\n category: ai-ml\n date: 2026-01-14\n heroImage: https://res.cloudinary.com/about-gitlab-com/image/upload/v1765809212/noh0mdfn9o94ry9ykura.png\n authors:\n - Itzik Gan Baruch\n description: Deep dive into GitLab Duo Agent Platform flows. Learn about foundational flows, create custom user-defined workflows, and learn flow orchestration patterns.\n title: 'Understanding flows: Multi-agent workflows'\n body: >-\n\n *Welcome to Part 4 of our eight-part guide, [Getting started with GitLab Duo Agent Platform](/blog/gitlab-duo-agent-platform-complete-getting-started-guide/), where you'll master building and deploying AI agents and workflows within your development lifecycle. Follow tutorials that take you from your first interaction to production-ready automation workflows with full customization.*\n\n\n **In this article:**\n\n\n - [What are flows and how do they work?](#introduction-to-flows)\n\n - [Foundational flows provided by GitLab](#foundational-flows)\n\n - [Creating custom flows](#how-to-create-custom-flows)\n\n - [Flow execution and orchestration](#flow-execution)\n\n - [Real-world examples and use cases](#example-custom-flow-yaml)\n\n\n > 🎯 Try [**GitLab Duo Agent Platform**](https://about.gitlab.com/gitlab-duo-agent-platform/) today!\n\n ## Introduction to flows\n\n Flows are combinations of one or more agents collaborating together. They orchestrate multi-step workflows to solve complex problems, and are executed on the GitLab platform compute.\n\n\n **Key characteristics of flows:**\n\n\n - **Multi-agent orchestration**: Combine multiple specialized agents\n\n - **Built-in**: Run on platform compute, no extra environment necessary\n\n - **Event-driven**: Triggered by mention, assignment, or assign as reviewer\n\n - **Asynchronous**: Run in background while you continue working\n\n - **Complete workflows**: Handle end-to-end tasks from analysis to implementation\n\n\n Think of flows as autonomous workflows that can gather context, make decisions, execute changes, and deliver results, all while you focus on other work.\n\n\n ## Flows vs. agents: Understanding the difference\n\n Agents work with you interactively. Flows work for you autonomously.\n\n\n | Aspect | Agents | Flows |\n\n |--------|--------|-------|\n\n | **Interaction** | Interactive chat | Autonomous execution |\n\n | **When to use** | Questions, guidance, and performing tasks interactively | Autonomous multi-step workflows |\n\n | **User involvement** | Active conversation | Trigger and review results |\n\n | **Execution time** | Real-time responses | Background processing |\n\n | **Complexity** | Single-agent tasks | Multi-agent orchestration |\n\n\n ## Flow types overview\n\n\n | Type | Interface | Maintainer | Use Case |\n\n |------|-------------|---------------|----------|\n\n | **[Foundational](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/)** | UI actions, IDE interface | GitLab | Software Development, Developer in issues, Fix CI/CD Pipeline, Convert to GitLab CI/CD, Code Review, SAST false positive detection |\n\n | **[Custom](https://docs.gitlab.com/user/duo_agent_platform/flows/custom/)** | Mention, assign, assign reviewer | You | Examples: Larger migration/modernization, release automation, dependency update management |\n\n\n ## Foundational flows\n\n Foundational flows are production-ready workflows created and maintained by GitLab. They're accessible through dedicated UI controls or IDE interfaces.\n\n ### Currently available foundational flows\n\n\n | Flow | Where Available | How to Access | Best For |\n\n |------|-----------------|---------------|----------|\n\n | [**Software Development**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/software_development/) | IDEs (VS Code, JetBrains, Visual Studio) | Flows tab in IDE | Feature implementation, complex refactoring, multi-file changes |\n\n | [**Developer**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/developer/) | GitLab Web UI | \"Generate MR with Duo\" button on issues | Well-defined features, bug fixes with clear steps |\n\n | [**Fix CI/CD Pipeline**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/fix_pipeline/) | GitLab Web UI | Failed pipeline interface | Pipeline debugging, CI/CD configuration issues |\n\n | [**Convert to GitLab CI/CD**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/convert_to_gitlab_ci/) | GitLab Web UI | \"Convert to GitLab CI/CD\" button on Jenkinsfile | Jenkins to GitLab CI/CD migration |\n\n | [**Code Review**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/code_review/) | GitLab Web UI | Assign as reviewer on MR | Automated code review with AI-native analysis and feedback |\n\n | [**SAST false positive detection**](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/sast_false_positive_detection/) | GitLab Web UI | Security scan results | Automatically identify and filter false positives in SAST findings |\n\n\n ## Custom flows\n\n\n Custom flows are YAML-defined workflows you create for your team's specific needs. They run in GitLab Runner and can be triggered by GitLab events.\n\n\n > **🎯 Try it now:** [Interactive demo of Custom Flows](https://gitlab.navattic.com/custom-flows) β€” Explore how to create and configure Custom Flows.\n\n\n ### Why create custom flows?\n\n\n Custom flows automate repetitive multi-step tasks that are specific to your team's workflow. Unlike foundational flows that serve general purposes, custom flows are tailored to your organization's processes, tools, and requirements.\n\n\n **Common use cases:**\n\n\n - **Automated code review**: Multi-stage review process (security scan β†’ quality check β†’ style validation)\n\n - **Compliance checking**: Verify regulatory requirements, license compliance, or security policies on each MR\n\n - **Documentation generation**: Auto-update API docs, README files, or changelogs based on code changes\n\n - **Dependency management**: Weekly security scans, automated updates, and vulnerability reports\n\n - **Custom testing**: Specialized test suites for your tech stack or integration tests\n\n\n ### Real-world example\n\n\n A fintech company creates a compliance flow that runs on every merge request. When triggered by `@compliance-flow`, the flow executes the following steps:\n\n 1. **Security agent** scans code for PCI-DSS violations and checks for exposed sensitive data.\n\n 2. **Code review agent** verifies that changes follow secure coding standards and best practices.\n\n 3. **Documentation agent** checks that API changes include updated documentation.\n\n 4. **Summary agent** aggregates findings and posts a compliance report with pass/fail status.\n\n\n The entire compliance review happens automatically in 5-10 minutes, providing consistent checks across all merge requests.\n\n\n ### How to trigger custom flows\n\n\n Custom flows can be triggered in multiple ways:\n\n\n **1. Via mentions in Issues/MRs:**\n\n Mention the flow in a comment to trigger it. Example for a documentation generation flow:\n\n ```text\n\n @doc-generator Generate API documentation for this feature\n\n ```\n\n\n **2. By assigning the flow to an issue or MR:**\n\n Assign the flow using either:\n\n - **GitLab UI**: Click the \"Assign\" button on the issue/MR and select the flow\n\n - **Command**: Use the `/assign` command in a comment. Example:\n\n ```shell\n\n /assign @doc-generator\n\n ```\n\n\n **3. By assigning the flow as a reviewer:**\n\n Assign the flow as a reviewer on a merge request using either:\n\n - **GitLab UI**: Click the \"Assign reviewer\" button on the merge request and select the flow\n\n - **Command**: Use the `/assign_reviewer` command in a comment. Example:\n\n ```shell\n\n /assign_reviewer @doc-reviewer\n\n ```\n\n Any of these methods automatically triggers the flow to execute and perform its tasks.\n\n\n ### How to create custom flows\n\n\n Custom flows are created through the GitLab UI at **Automate β†’ Flows β†’ New flow** in your project, or from **Explore β†’ AI Catalog β†’ Flows β†’ New flow**. You define your flow using YAML configuration that specifies components, prompts, routing, and execution flow. The YAML schema allows you to create sophisticated multi-agent workflows with precise control over agent behavior and orchestration.\n\n\n\n **Key elements of a custom flow:**\n\n\n - **Components**: Define the agents and steps in your workflow\n\n - **Prompts**: Configure AI model behavior and instructions\n\n - **Routers**: Control the flow between components\n\n - **Toolsets**: Specify which GitLab API tools agents can use\n\n\n ### Example custom flow YAML\n\n **Background:** This example shows a feature implementation flow for a travel booking platform. When a developer creates an issue with feature requirements, they can trigger this flow to automatically analyze the requirements, review the codebase, implement the solution, and create a merge request, all without manual intervention.\n\n Here's the YAML configuration:\n\n\n ```yaml\n\n version: \"v1\"\n\n environment: ambient\n\n components:\n - name: \"implement_feature\"\n type: AgentComponent\n prompt_id: \"implementation_prompt\"\n inputs:\n - from: \"context:goal\"\n as: \"user_goal\"\n - from: \"context:project_id\"\n as: \"project_id\"\n toolset:\n - \"get_issue\"\n - \"get_repository_file\"\n - \"list_repository_tree\"\n - \"find_files\"\n - \"blob_search\"\n - \"create_file\"\n - \"create_commit\"\n - \"create_merge_request\"\n - \"create_issue_note\"\n ui_log_events:\n - \"on_agent_final_answer\"\n - \"on_tool_execution_success\"\n - \"on_tool_execution_failed\"\n\n prompts:\n - name: \"Cheapflights Feature Implementation\"\n prompt_id: \"implementation_prompt\"\n unit_primitives: []\n prompt_template:\n system: |\n You are an expert full-stack developer specializing in travel booking platforms, specifically Cheapflights.\n\n Your task is to:\n 1. Extract the issue IID from the goal (look for \"Issue IID: XX\")\n 2. Use get_issue with project_id={{project_id}} and issue_iid to retrieve issue details\n 3. Analyze the requirements for the flight search feature\n 4. Review the existing codebase using list_repository_tree, find_files, and get_repository_file\n 5. Design and implement the solution following Cheapflights best practices\n 6. Create all necessary code files using create_file (call multiple times for multiple files)\n 7. Commit the changes using create_commit\n 8. Create a merge request using create_merge_request\n 9. Post a summary comment to the issue using create_issue_note with the MR link\n\n Cheapflights Domain Expertise:\n - Flight search and booking systems (Amadeus, Sabre, Skyscanner APIs)\n - Fare comparison and pricing strategies\n - Real-time availability and inventory management\n - Travel industry UX patterns\n - Performance optimization for high-traffic flight searches\n\n Code Standards:\n - Clean, maintainable code (TypeScript/JavaScript/Python/React)\n - Proper state management for React components\n - RESTful API endpoints with comprehensive error handling\n - Mobile-first responsive design\n - Proper timezone handling (use moment-timezone or date-fns-tz)\n - WCAG 2.1 accessibility compliance\n\n Flight-Specific Best Practices:\n - Accurate fare calculations (base fare + taxes + fees + surcharges)\n - Flight duration calculations across timezones\n - Search filter logic (price range, number of stops, airlines, departure/arrival times)\n - Sort algorithms (best value, fastest, cheapest)\n - Handle edge cases: date line crossing, daylight saving time, red-eye flights\n - Currency amounts use proper decimal handling (avoid floating point errors)\n - Dates use ISO 8601 format\n - Flight codes follow IATA standards (3-letter airport codes)\n\n Implementation Requirements:\n - No TODOs or placeholder comments\n - All functions must be fully implemented\n - Include proper TypeScript types or Python type hints\n - Add JSDoc/docstring comments for all functions\n - Comprehensive error handling and input validation\n - Basic unit tests for critical functions\n - Performance considerations for handling large result sets\n\n CRITICAL - Your final comment on the issue MUST include:\n - **Implementation Summary**: Brief description of what was implemented\n - **Files Created/Modified**: List of all files with descriptions\n - **Key Features**: Bullet points of main functionality\n - **Technical Approach**: Brief explanation of architecture/patterns used\n - **Testing Notes**: How to test the implementation\n - **Merge Request Link**: Direct link to the created MR (format: [View Merge Request](MR_URL))\n\n IMPORTANT TOOL USAGE:\n - Extract the issue IID from the goal first (e.g., \"Issue IID: 12\" means issue_iid=12)\n - Use get_issue with project_id={{project_id}} and issue_iid=\u003Cextracted_iid>\n - Create multiple files by calling create_file multiple times (once per file)\n - Use create_commit to commit all files together with a descriptive commit message\n - Use create_merge_request to create the MR and capture the MR URL from the response\n - Use create_issue_note with project_id={{project_id}}, noteable_id=\u003Cissue_iid>, and body=\u003Cyour complete summary with MR link>\n - Make sure to include the MR link in the comment body so users can easily access it\n\n user: |\n Goal: {{user_goal}}\n Project ID: {{project_id}}\n\n Please complete the following steps:\n 1. Extract the issue IID and retrieve full issue details\n 2. Analyze the requirements thoroughly\n 3. Review the existing codebase structure and patterns\n 4. Implement the feature with production-ready code\n 5. Create all necessary files (components, APIs, tests, documentation)\n 6. Commit all changes with a clear commit message\n 7. Create a merge request\n 8. Post a detailed summary comment to the issue including the MR link\n\n placeholder: history\n params:\n timeout: 300\n\n routers:\n - from: \"implement_feature\"\n to: \"end\"\n\n flow:\n entry_point: \"implement_feature\"\n ```\n\n **What this flow does:** This flow orchestrates an AI agent to automatically implement a feature by analyzing issue requirements, reviewing the codebase, writing production-ready code with domain expertise, and creating a merge request with a detailed summary comment.\n\n\n For complete documentation and examples, see:\n\n\n - [Custom Flows documentation](https://docs.gitlab.com/user/duo_agent_platform/flows/custom/)\n\n - [Flow Registry Framework (YAML Schema)](https://gitlab.com/gitlab-org/modelops/applied-ml/code-suggestions/ai-assist/-/blob/main/docs/flow_registry/v1.md)\n\n\n ## Flow execution\n\n\n Flows run on GitLab platform compute. When triggered by an event (mention, assignment, or button click), a session is created and the flow starts to execute.\n\n ### Available environment variables\n\n\n Flows have access to environment variables that provide context about the trigger and the GitLab object:\n\n\n - **`AI_FLOW_CONTEXT`** β€” JSON-serialized context including MR diffs, issue descriptions, comments, and discussion threads\n\n - **`AI_FLOW_INPUT`** β€” The user's prompt or comment text that triggered the flow\n\n - **`AI_FLOW_EVENT`** β€” The event type that triggered the flow (`mention`, `assign`, `assign_reviewer`)\n\n\n These variables allow your flow to understand what triggered it and access the relevant GitLab data to perform its task.\n\n\n ### Multi-agent flows\n\n\n Custom flows can include multiple agent components that work together sequentially. The flow's YAML configuration defines:\n\n\n - **Components**: One or more agents (AgentComponent) or deterministic steps\n\n - **Routers**: Define the flow between components (e.g., from component A to component B to end)\n\n - **Prompts**: Configure each agent's behavior and model\n\n\n For example, a code review flow might have a security agent, then a quality agent, then an approval agent, with routers connecting them in sequence.\n\n\n ### Monitoring flow execution\n\n\n To view flows that are running for your project:\n\n\n 1. Navigate to **Automate β†’ Sessions**.\n\n 2. Select any session to view more details.\n\n 3. The **Details** tab shows a link to the CI/CD job logs.\n\n\n Sessions show detailed information including step-by-step progress, tool invocations, reasoning, and decision-making process.\n\n ### When to use flows\n\n - Complex multi-step tasks\n\n - Background automation\n\n - Event-driven workflows\n\n - Multi-file changes\n\n - Tasks that take time\n\n - Automated reviews/checks\n\n ## What's next?\n\n You now understand flows, how to create them, and when to use them vs. agents. Next, learn how to discover, create, and share agents and flows across your organization in [Part 5: AI Catalog](/blog/ai-catalog-discover-and-share-agents/). Explore the AI Catalog to browse available flows and agents, add them to your projects, and publish your own agents and flows.\n\n\n ## Resources\n\n\n - [GitLab Duo Agent Platform Flows](https://docs.gitlab.com/user/duo_agent_platform/flows/)\n\n - [Foundational Flows documentation](https://docs.gitlab.com/user/duo_agent_platform/flows/foundational_flows/)\n\n - [Custom Flows documentation](https://docs.gitlab.com/user/duo_agent_platform/flows/custom/)\n\n - [Flow execution configuration](https://docs.gitlab.com/user/duo_agent_platform/flows/execution/)\n\n - [GitLab CI/CD Variables guide](https://docs.gitlab.com/ci/variables/)\n\n - [Service Accounts](https://docs.gitlab.com/user/profile/service_accounts/)\n\n\n ---\n\n **Next:** [Part 5: AI Catalog](/blog/ai-catalog-discover-and-share-agents/)\n\n\n **Previous:** [Part 3: Understanding agents](/blog/understanding-agents-foundational-custom-external/)\n\nconfig:\n slug: understanding-flows-multi-agent-workflows\n featured: false\n template: BlogPost\n",{"noIndex":14,"title":5,"description":25,"ogImage":33},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1752678395/impw8no5tbskr6k2afgu.jpg","en-us/blog/understanding-flows-multi-agent-workflows",[36,19,20,21],"aiml",[18,19,20,21],"aFGiN9Bu8GkVLMpmIxCbV8X_B72Bi6TuvVQRrF2e4fs",{"data":40},{"logo":41,"freeTrial":46,"sales":51,"login":56,"items":61,"search":370,"minimal":401,"duo":420,"switchNav":429,"pricingDeployment":440},{"config":42},{"href":43,"dataGaName":44,"dataGaLocation":45},"/","gitlab logo","header",{"text":47,"config":48},"Get free trial",{"href":49,"dataGaName":50,"dataGaLocation":45},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com&glm_content=default-saas-trial/","free trial",{"text":52,"config":53},"Talk to sales",{"href":54,"dataGaName":55,"dataGaLocation":45},"/sales/","sales",{"text":57,"config":58},"Sign in",{"href":59,"dataGaName":60,"dataGaLocation":45},"https://gitlab.com/users/sign_in/","sign in",[62,89,184,189,291,351],{"text":63,"config":64,"cards":66},"Platform",{"dataNavLevelOne":65},"platform",[67,73,81],{"title":63,"description":68,"link":69},"The intelligent orchestration platform for DevSecOps",{"text":70,"config":71},"Explore our Platform",{"href":72,"dataGaName":65,"dataGaLocation":45},"/platform/",{"title":74,"description":75,"link":76},"GitLab Duo Agent Platform","Agentic AI for the entire software lifecycle",{"text":77,"config":78},"Meet GitLab Duo",{"href":79,"dataGaName":80,"dataGaLocation":45},"/gitlab-duo-agent-platform/","gitlab duo agent platform",{"title":82,"description":83,"link":84},"Why GitLab","See the top reasons enterprises choose GitLab",{"text":85,"config":86},"Learn more",{"href":87,"dataGaName":88,"dataGaLocation":45},"/why-gitlab/","why gitlab",{"text":90,"left":29,"config":91,"link":93,"lists":97,"footer":166},"Product",{"dataNavLevelOne":92},"solutions",{"text":94,"config":95},"View all Solutions",{"href":96,"dataGaName":92,"dataGaLocation":45},"/solutions/",[98,122,145],{"title":99,"description":100,"link":101,"items":106},"Automation","CI/CD and automation to accelerate deployment",{"config":102},{"icon":103,"href":104,"dataGaName":105,"dataGaLocation":45},"AutomatedCodeAlt","/solutions/delivery-automation/","automated software delivery",[107,111,114,118],{"text":108,"config":109},"CI/CD",{"href":110,"dataGaLocation":45,"dataGaName":108},"/solutions/continuous-integration/",{"text":74,"config":112},{"href":79,"dataGaLocation":45,"dataGaName":113},"gitlab duo agent platform - product menu",{"text":115,"config":116},"Source Code Management",{"href":117,"dataGaLocation":45,"dataGaName":115},"/solutions/source-code-management/",{"text":119,"config":120},"Automated Software Delivery",{"href":104,"dataGaLocation":45,"dataGaName":121},"Automated software delivery",{"title":123,"description":124,"link":125,"items":130},"Security","Deliver code faster without compromising security",{"config":126},{"href":127,"dataGaName":128,"dataGaLocation":45,"icon":129},"/solutions/application-security-testing/","security and compliance","ShieldCheckLight",[131,135,140],{"text":132,"config":133},"Application Security Testing",{"href":127,"dataGaName":134,"dataGaLocation":45},"Application security testing",{"text":136,"config":137},"Software Supply Chain Security",{"href":138,"dataGaLocation":45,"dataGaName":139},"/solutions/supply-chain/","Software supply chain security",{"text":141,"config":142},"Software Compliance",{"href":143,"dataGaName":144,"dataGaLocation":45},"/solutions/software-compliance/","software compliance",{"title":146,"link":147,"items":152},"Measurement",{"config":148},{"icon":149,"href":150,"dataGaName":151,"dataGaLocation":45},"DigitalTransformation","/solutions/visibility-measurement/","visibility and measurement",[153,157,161],{"text":154,"config":155},"Visibility & Measurement",{"href":150,"dataGaLocation":45,"dataGaName":156},"Visibility and Measurement",{"text":158,"config":159},"Value Stream Management",{"href":160,"dataGaLocation":45,"dataGaName":158},"/solutions/value-stream-management/",{"text":162,"config":163},"Analytics & Insights",{"href":164,"dataGaLocation":45,"dataGaName":165},"/solutions/analytics-and-insights/","Analytics and insights",{"title":167,"items":168},"GitLab for",[169,174,179],{"text":170,"config":171},"Enterprise",{"href":172,"dataGaLocation":45,"dataGaName":173},"/enterprise/","enterprise",{"text":175,"config":176},"Small Business",{"href":177,"dataGaLocation":45,"dataGaName":178},"/small-business/","small business",{"text":180,"config":181},"Public Sector",{"href":182,"dataGaLocation":45,"dataGaName":183},"/solutions/public-sector/","public sector",{"text":185,"config":186},"Pricing",{"href":187,"dataGaName":188,"dataGaLocation":45,"dataNavLevelOne":188},"/pricing/","pricing",{"text":190,"config":191,"link":193,"lists":197,"feature":282},"Resources",{"dataNavLevelOne":192},"resources",{"text":194,"config":195},"View all resources",{"href":196,"dataGaName":192,"dataGaLocation":45},"/resources/",[198,231,254],{"title":199,"items":200},"Getting started",[201,206,211,216,221,226],{"text":202,"config":203},"Install",{"href":204,"dataGaName":205,"dataGaLocation":45},"/install/","install",{"text":207,"config":208},"Quick start guides",{"href":209,"dataGaName":210,"dataGaLocation":45},"/get-started/","quick setup checklists",{"text":212,"config":213},"Learn",{"href":214,"dataGaLocation":45,"dataGaName":215},"https://university.gitlab.com/","learn",{"text":217,"config":218},"Product documentation",{"href":219,"dataGaName":220,"dataGaLocation":45},"https://docs.gitlab.com/","product documentation",{"text":222,"config":223},"Best practice videos",{"href":224,"dataGaName":225,"dataGaLocation":45},"/getting-started-videos/","best practice videos",{"text":227,"config":228},"Integrations",{"href":229,"dataGaName":230,"dataGaLocation":45},"/integrations/","integrations",{"title":232,"items":233},"Discover",[234,239,244,249],{"text":235,"config":236},"Customer success stories",{"href":237,"dataGaName":238,"dataGaLocation":45},"/customers/","customer success stories",{"text":240,"config":241},"Blog",{"href":242,"dataGaName":243,"dataGaLocation":45},"/blog/","blog",{"text":245,"config":246},"The Source",{"href":247,"dataGaName":248,"dataGaLocation":45},"/the-source/","the source",{"text":250,"config":251},"Remote",{"href":252,"dataGaName":253,"dataGaLocation":45},"https://handbook.gitlab.com/handbook/company/culture/all-remote/","remote",{"title":255,"items":256},"Connect",[257,262,267,272,277],{"text":258,"config":259},"GitLab Services",{"href":260,"dataGaName":261,"dataGaLocation":45},"/services/","services",{"text":263,"config":264},"Community",{"href":265,"dataGaName":266,"dataGaLocation":45},"/community/","community",{"text":268,"config":269},"Forum",{"href":270,"dataGaName":271,"dataGaLocation":45},"https://forum.gitlab.com/","forum",{"text":273,"config":274},"Events",{"href":275,"dataGaName":276,"dataGaLocation":45},"/events/","events",{"text":278,"config":279},"Partners",{"href":280,"dataGaName":281,"dataGaLocation":45},"/partners/","partners",{"textColor":283,"title":284,"text":285,"link":286},"#000","What’s new in GitLab","Stay updated with our latest features and improvements.",{"text":287,"config":288},"Read the latest",{"href":289,"dataGaName":290,"dataGaLocation":45},"/releases/whats-new/","whats new",{"text":292,"config":293,"lists":295},"Company",{"dataNavLevelOne":294},"company",[296],{"items":297},[298,303,309,311,316,321,326,331,336,341,346],{"text":299,"config":300},"About",{"href":301,"dataGaName":302,"dataGaLocation":45},"/company/","about",{"text":304,"config":305,"footerGa":308},"Jobs",{"href":306,"dataGaName":307,"dataGaLocation":45},"/jobs/","jobs",{"dataGaName":307},{"text":273,"config":310},{"href":275,"dataGaName":276,"dataGaLocation":45},{"text":312,"config":313},"Leadership",{"href":314,"dataGaName":315,"dataGaLocation":45},"/company/team/e-group/","leadership",{"text":317,"config":318},"Team",{"href":319,"dataGaName":320,"dataGaLocation":45},"/company/team/","team",{"text":322,"config":323},"Handbook",{"href":324,"dataGaName":325,"dataGaLocation":45},"https://handbook.gitlab.com/","handbook",{"text":327,"config":328},"Investor relations",{"href":329,"dataGaName":330,"dataGaLocation":45},"https://ir.gitlab.com/","investor relations",{"text":332,"config":333},"Trust Center",{"href":334,"dataGaName":335,"dataGaLocation":45},"/security/","trust center",{"text":337,"config":338},"AI Transparency Center",{"href":339,"dataGaName":340,"dataGaLocation":45},"/ai-transparency-center/","ai transparency center",{"text":342,"config":343},"Newsletter",{"href":344,"dataGaName":345,"dataGaLocation":45},"/company/contact/#contact-forms","newsletter",{"text":347,"config":348},"Press",{"href":349,"dataGaName":350,"dataGaLocation":45},"/press/","press",{"text":352,"config":353,"lists":354},"Contact us",{"dataNavLevelOne":294},[355],{"items":356},[357,360,365],{"text":52,"config":358},{"href":54,"dataGaName":359,"dataGaLocation":45},"talk to sales",{"text":361,"config":362},"Support portal",{"href":363,"dataGaName":364,"dataGaLocation":45},"https://support.gitlab.com","support portal",{"text":366,"config":367},"Customer portal",{"href":368,"dataGaName":369,"dataGaLocation":45},"https://customers.gitlab.com/customers/sign_in/","customer portal",{"close":371,"login":372,"suggestions":379},"Close",{"text":373,"link":374},"To search repositories and projects, login to",{"text":375,"config":376},"gitlab.com",{"href":59,"dataGaName":377,"dataGaLocation":378},"search login","search",{"text":380,"default":381},"Suggestions",[382,384,388,390,394,398],{"text":74,"config":383},{"href":79,"dataGaName":74,"dataGaLocation":378},{"text":385,"config":386},"Code Suggestions (AI)",{"href":387,"dataGaName":385,"dataGaLocation":378},"/solutions/code-suggestions/",{"text":108,"config":389},{"href":110,"dataGaName":108,"dataGaLocation":378},{"text":391,"config":392},"GitLab on AWS",{"href":393,"dataGaName":391,"dataGaLocation":378},"/partners/technology-partners/aws/",{"text":395,"config":396},"GitLab on Google Cloud",{"href":397,"dataGaName":395,"dataGaLocation":378},"/partners/technology-partners/google-cloud-platform/",{"text":399,"config":400},"Why GitLab?",{"href":87,"dataGaName":399,"dataGaLocation":378},{"freeTrial":402,"mobileIcon":407,"desktopIcon":412,"secondaryButton":415},{"text":403,"config":404},"Start free trial",{"href":405,"dataGaName":50,"dataGaLocation":406},"https://gitlab.com/-/trials/new/","nav",{"altText":408,"config":409},"Gitlab Icon",{"src":410,"dataGaName":411,"dataGaLocation":406},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758203874/jypbw1jx72aexsoohd7x.svg","gitlab icon",{"altText":408,"config":413},{"src":414,"dataGaName":411,"dataGaLocation":406},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1758203875/gs4c8p8opsgvflgkswz9.svg",{"text":416,"config":417},"Get Started",{"href":418,"dataGaName":419,"dataGaLocation":406},"https://gitlab.com/-/trial_registrations/new?glm_source=about.gitlab.com/get-started/","get started",{"freeTrial":421,"mobileIcon":425,"desktopIcon":427},{"text":422,"config":423},"Learn more about GitLab Duo",{"href":79,"dataGaName":424,"dataGaLocation":406},"gitlab duo",{"altText":408,"config":426},{"src":410,"dataGaName":411,"dataGaLocation":406},{"altText":408,"config":428},{"src":414,"dataGaName":411,"dataGaLocation":406},{"button":430,"mobileIcon":435,"desktopIcon":437},{"text":431,"config":432},"/switch",{"href":433,"dataGaName":434,"dataGaLocation":406},"#contact","switch",{"altText":408,"config":436},{"src":410,"dataGaName":411,"dataGaLocation":406},{"altText":408,"config":438},{"src":439,"dataGaName":411,"dataGaLocation":406},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1773335277/ohhpiuoxoldryzrnhfrh.png",{"freeTrial":441,"mobileIcon":446,"desktopIcon":448},{"text":442,"config":443},"Back to pricing",{"href":187,"dataGaName":444,"dataGaLocation":406,"icon":445},"back to pricing","GoBack",{"altText":408,"config":447},{"src":410,"dataGaName":411,"dataGaLocation":406},{"altText":408,"config":449},{"src":414,"dataGaName":411,"dataGaLocation":406},{"title":451,"button":452,"config":457},"See how agentic AI transforms software delivery",{"text":453,"config":454},"Watch GitLab Transcend now",{"href":455,"dataGaName":456,"dataGaLocation":45},"/events/transcend/virtual/","transcend event",{"layout":458,"icon":459,"disabled":29},"release","AiStar",{"data":461},{"text":462,"source":463,"edit":469,"contribute":474,"config":479,"items":484,"minimal":691},"Git is a trademark of Software Freedom Conservancy and our use of 'GitLab' is under license",{"text":464,"config":465},"View page source",{"href":466,"dataGaName":467,"dataGaLocation":468},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/","page source","footer",{"text":470,"config":471},"Edit this page",{"href":472,"dataGaName":473,"dataGaLocation":468},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/content/","web ide",{"text":475,"config":476},"Please contribute",{"href":477,"dataGaName":478,"dataGaLocation":468},"https://gitlab.com/gitlab-com/marketing/digital-experience/about-gitlab-com/-/blob/main/CONTRIBUTING.md/","please contribute",{"twitter":480,"facebook":481,"youtube":482,"linkedin":483},"https://twitter.com/gitlab","https://www.facebook.com/gitlab","https://www.youtube.com/channel/UCnMGQ8QHMAnVIsI3xJrihhg","https://www.linkedin.com/company/gitlab-com",[485,532,586,630,657],{"title":185,"links":486,"subMenu":501},[487,491,496],{"text":488,"config":489},"View plans",{"href":187,"dataGaName":490,"dataGaLocation":468},"view plans",{"text":492,"config":493},"Why Premium?",{"href":494,"dataGaName":495,"dataGaLocation":468},"/pricing/premium/","why premium",{"text":497,"config":498},"Why Ultimate?",{"href":499,"dataGaName":500,"dataGaLocation":468},"/pricing/ultimate/","why ultimate",[502],{"title":503,"links":504},"Contact Us",[505,508,510,512,517,522,527],{"text":506,"config":507},"Contact sales",{"href":54,"dataGaName":55,"dataGaLocation":468},{"text":361,"config":509},{"href":363,"dataGaName":364,"dataGaLocation":468},{"text":366,"config":511},{"href":368,"dataGaName":369,"dataGaLocation":468},{"text":513,"config":514},"Status",{"href":515,"dataGaName":516,"dataGaLocation":468},"https://status.gitlab.com/","status",{"text":518,"config":519},"Terms of use",{"href":520,"dataGaName":521,"dataGaLocation":468},"/terms/","terms of use",{"text":523,"config":524},"Privacy statement",{"href":525,"dataGaName":526,"dataGaLocation":468},"/privacy/","privacy statement",{"text":528,"config":529},"Cookie preferences",{"dataGaName":530,"dataGaLocation":468,"id":531,"isOneTrustButton":29},"cookie preferences","ot-sdk-btn",{"title":90,"links":533,"subMenu":542},[534,538],{"text":535,"config":536},"DevSecOps platform",{"href":72,"dataGaName":537,"dataGaLocation":468},"devsecops platform",{"text":539,"config":540},"AI-Assisted Development",{"href":79,"dataGaName":541,"dataGaLocation":468},"ai-assisted development",[543],{"title":544,"links":545},"Topics",[546,551,556,561,566,571,576,581],{"text":547,"config":548},"CICD",{"href":549,"dataGaName":550,"dataGaLocation":468},"/topics/ci-cd/","cicd",{"text":552,"config":553},"GitOps",{"href":554,"dataGaName":555,"dataGaLocation":468},"/topics/gitops/","gitops",{"text":557,"config":558},"DevOps",{"href":559,"dataGaName":560,"dataGaLocation":468},"/topics/devops/","devops",{"text":562,"config":563},"Version Control",{"href":564,"dataGaName":565,"dataGaLocation":468},"/topics/version-control/","version control",{"text":567,"config":568},"DevSecOps",{"href":569,"dataGaName":570,"dataGaLocation":468},"/topics/devsecops/","devsecops",{"text":572,"config":573},"Cloud Native",{"href":574,"dataGaName":575,"dataGaLocation":468},"/topics/cloud-native/","cloud native",{"text":577,"config":578},"AI for Coding",{"href":579,"dataGaName":580,"dataGaLocation":468},"/topics/devops/ai-for-coding/","ai for coding",{"text":582,"config":583},"Agentic AI",{"href":584,"dataGaName":585,"dataGaLocation":468},"/topics/agentic-ai/","agentic ai",{"title":587,"links":588},"Solutions",[589,591,593,598,602,605,609,612,614,617,620,625],{"text":132,"config":590},{"href":127,"dataGaName":132,"dataGaLocation":468},{"text":121,"config":592},{"href":104,"dataGaName":105,"dataGaLocation":468},{"text":594,"config":595},"Agile development",{"href":596,"dataGaName":597,"dataGaLocation":468},"/solutions/agile-delivery/","agile delivery",{"text":599,"config":600},"SCM",{"href":117,"dataGaName":601,"dataGaLocation":468},"source code management",{"text":547,"config":603},{"href":110,"dataGaName":604,"dataGaLocation":468},"continuous integration & delivery",{"text":606,"config":607},"Value stream management",{"href":160,"dataGaName":608,"dataGaLocation":468},"value stream management",{"text":552,"config":610},{"href":611,"dataGaName":555,"dataGaLocation":468},"/solutions/gitops/",{"text":170,"config":613},{"href":172,"dataGaName":173,"dataGaLocation":468},{"text":615,"config":616},"Small business",{"href":177,"dataGaName":178,"dataGaLocation":468},{"text":618,"config":619},"Public sector",{"href":182,"dataGaName":183,"dataGaLocation":468},{"text":621,"config":622},"Education",{"href":623,"dataGaName":624,"dataGaLocation":468},"/solutions/education/","education",{"text":626,"config":627},"Financial services",{"href":628,"dataGaName":629,"dataGaLocation":468},"/solutions/finance/","financial services",{"title":190,"links":631},[632,634,636,638,641,643,645,647,649,651,653,655],{"text":202,"config":633},{"href":204,"dataGaName":205,"dataGaLocation":468},{"text":207,"config":635},{"href":209,"dataGaName":210,"dataGaLocation":468},{"text":212,"config":637},{"href":214,"dataGaName":215,"dataGaLocation":468},{"text":217,"config":639},{"href":219,"dataGaName":640,"dataGaLocation":468},"docs",{"text":240,"config":642},{"href":242,"dataGaName":243,"dataGaLocation":468},{"text":235,"config":644},{"href":237,"dataGaName":238,"dataGaLocation":468},{"text":250,"config":646},{"href":252,"dataGaName":253,"dataGaLocation":468},{"text":258,"config":648},{"href":260,"dataGaName":261,"dataGaLocation":468},{"text":263,"config":650},{"href":265,"dataGaName":266,"dataGaLocation":468},{"text":268,"config":652},{"href":270,"dataGaName":271,"dataGaLocation":468},{"text":273,"config":654},{"href":275,"dataGaName":276,"dataGaLocation":468},{"text":278,"config":656},{"href":280,"dataGaName":281,"dataGaLocation":468},{"title":292,"links":658},[659,661,663,665,667,669,671,675,680,682,684,686],{"text":299,"config":660},{"href":301,"dataGaName":294,"dataGaLocation":468},{"text":304,"config":662},{"href":306,"dataGaName":307,"dataGaLocation":468},{"text":312,"config":664},{"href":314,"dataGaName":315,"dataGaLocation":468},{"text":317,"config":666},{"href":319,"dataGaName":320,"dataGaLocation":468},{"text":322,"config":668},{"href":324,"dataGaName":325,"dataGaLocation":468},{"text":327,"config":670},{"href":329,"dataGaName":330,"dataGaLocation":468},{"text":672,"config":673},"Sustainability",{"href":674,"dataGaName":672,"dataGaLocation":468},"/sustainability/",{"text":676,"config":677},"Diversity, inclusion and belonging (DIB)",{"href":678,"dataGaName":679,"dataGaLocation":468},"/diversity-inclusion-belonging/","Diversity, inclusion and belonging",{"text":332,"config":681},{"href":334,"dataGaName":335,"dataGaLocation":468},{"text":342,"config":683},{"href":344,"dataGaName":345,"dataGaLocation":468},{"text":347,"config":685},{"href":349,"dataGaName":350,"dataGaLocation":468},{"text":687,"config":688},"Modern Slavery Transparency Statement",{"href":689,"dataGaName":690,"dataGaLocation":468},"https://handbook.gitlab.com/handbook/legal/modern-slavery-act-transparency-statement/","modern slavery transparency statement",{"items":692},[693,696,699],{"text":694,"config":695},"Terms",{"href":520,"dataGaName":521,"dataGaLocation":468},{"text":697,"config":698},"Cookies",{"dataGaName":530,"dataGaLocation":468,"id":531,"isOneTrustButton":29},{"text":700,"config":701},"Privacy",{"href":525,"dataGaName":526,"dataGaLocation":468},[703],{"id":704,"title":9,"body":27,"config":705,"content":707,"description":27,"extension":26,"meta":711,"navigation":29,"path":712,"seo":713,"stem":714,"__hash__":715},"blogAuthors/en-us/blog/authors/itzik-gan-baruch.yml",{"template":706},"BlogAuthor",{"name":9,"config":708},{"headshot":709,"ctfId":710},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1749658921/Blog/Author%20Headshots/iganbaruch-headshot.jpg","iganbaruch",{},"/en-us/blog/authors/itzik-gan-baruch",{},"en-us/blog/authors/itzik-gan-baruch","bz9VMiTQ1ixvnoxUFk0jiUcnLG3oQsymgXNCqyRqfsk",[717,729,741],{"content":718,"config":727},{"title":719,"description":720,"authors":721,"heroImage":723,"date":724,"body":725,"category":11,"tags":726},"GitLab and Anthropic: Governed AI for enterprise development","GitLab deepens its Anthropic Claude integration, bringing governed AI, access to new models, and cloud flexibility to enterprise software development.",[722],"Stuart Moncada","https://res.cloudinary.com/about-gitlab-com/image/upload/v1776457632/llddiylsgwuze0u1rjks.png","2026-04-28","For enterprise and public sector leaders, the tension is familiar: Software teams need to move faster with AI, while security, compliance, and regulatory expectations only get more stringent. GitLab deepens its Anthropic Claude integration so organizations get access to newly released Claude models inside GitLab’s intelligent orchestration platform where governance, compliance, and auditability already run.\n\nClaude powers capabilities across GitLab Duo Agent Platform as the default model out of the box, across a variety of use cases from code generation and review to agentic chat and vulnerability resolution. If you've used GitLab Duo, you've already experienced how Duo agents automate workflows across the entire software development lifecycle (SDLC).\n\nThis accelerates the integration of Claude’s capabilities into GitLab, broadens how enterprises can deploy them, and reinforces what makes GitLab fundamentally different as a platform for software development and engineering: governance, compliance, and auditability built into every AI interaction.\n\n> \"GitLab Duo has accelerated how our teams plan, build, and ship software. The combination of Anthropic's Claude and GitLab's platform means we're getting more capable AI without changing how we work or how it is governed.\"\n>\n> – Mans Booijink, Operations Manager, Cube\n\n## The real differentiator: Governed AI\n\nWith GitLab, governance controls and auditing are built into the SDLC. When Claude suggests a code change through the GitLab Duo Agent Platform, that suggestion flows through the same merge request process, the same approval rules, the same security scanning, and the same audit trail as every other change. AI doesn't get a shortcut around your controls. It operates within them.\n\nAs GitLab moves deeper into agentic software development, where AI autonomously handles well-defined tasks, the governance layer becomes more important. An AI agent that can open a merge request, help resolve a vulnerability, or refactor a service needs to be auditable, attributable, and subject to the same policy enforcement as a human developer. That requirement is an architectural decision GitLab made from the start, and one that grows more consequential as AI agents take on broader responsibilities.\n\n## Enterprise deployment flexibility\n\nThis also expands how organizations access the latest Claude models through GitLab. Claude is available within GitLab through Google Cloud's Vertex AI and Amazon Bedrock, which means enterprises can route AI workloads through the hyperscaler commitments and cloud governance frameworks they already have in place. No separate vendor contract. No new data residency questions. Your existing Google Cloud or AWS relationship is the on-ramp. \n\nGitLab is now also available in the [Claude Marketplace](https://claude.com/platform/marketplace), allowing customers to purchase GitLab Credits and apply them toward existing Anthropic spending commitments – consolidating AI spend and simplifying how teams discover and procure GitLab alongside their Anthropic investments.\n\n## Advancing an agentic future\n\nGitLab's vision for agentic software development, where AI handles defined tasks autonomously across planning, coding, testing, securing, and deploying, requires models with strong reasoning, reliability, and safety characteristics. It also requires a platform where those autonomous actions are fully governed.\n\nAgentic workflows demand models with strong reasoning, reliability, and safety characteristics, criteria that guide how GitLab selects and integrates AI model partners. And GitLab's governance framework helps ensure that as AI agents assume more advanced development work, enterprises maintain full visibility and control over what those agents do, when they do it, and how changes are tracked.\n\n## What this means for GitLab customers\n\nIf you're already using GitLab Duo Agent Platform, you'll get access to Claude models and deeper AI assistance across your software development lifecycle, all within the governance framework you already rely on.\n\nIf you're evaluating AI-powered software development platforms, you shouldn't have to choose between advanced AI capabilities and enterprise control. This strategic integration is built to deliver both.\n\n> Want to learn more about GitLab Duo Agent Platform? [Get a demo or start a free trial today](https://about.gitlab.com/gitlab-duo-agent-platform/).",[18,19,281],{"featured":29,"template":15,"slug":728},"gitlab-and-anthropic-governed-ai-for-enterprise-development",{"content":730,"config":739},{"title":731,"description":732,"authors":733,"heroImage":735,"date":736,"body":737,"category":11,"tags":738},"Give your AI agent direct, structured GitLab access with glab CLI","The GitLab CLI (glab) provides AI agents structured, reliable access to projects via the MCP, eliminating friction. This tutorial shows how you can speed up code review and issue triage.",[734],"Kai Armstrong","https://res.cloudinary.com/about-gitlab-com/image/upload/v1776347152/unw3mzatkd5xyfbzcnni.png","2026-04-27","\nWhen teams use GitLab Duo, Claude, Cursor, and other AI assistants, more of the development workflow runs through an AI agent acting on your behalf β€” reading issues, reviewing merge requests, running pipelines, and helping you ship faster. Most developers are already using the GitLab CLI (`glab`) from the terminal to interact with GitLab. Combining the two is a natural next step.\n\n\nThe problem is that without the right tools, AI agents are essentially guessing when it comes to your GitLab projects. They might hallucinate the details of an issue they've never seen, summarize a merge request based on stale training data rather than its actual state, or require you to manually copy context from a browser tab and paste it into a chat window just to get started. Every one of those workarounds is friction: it slows you down, introduces the possibility of error, and puts a hard ceiling on what your agent can actually do on your behalf. `glab` changes that by giving agents a direct, reliable interface to your projects.\n\n\nWith `glab`, your agent fetches what it needs directly from GitLab, acts on it, and reports back β€” so you spend less time relaying information and more time on the work that matters.\n\n\nIn this tutorial, you'll learn how to use `glab` to give AI agents structured, reliable access to your GitLab projects. You'll also discover how that unlocks a faster, more capable development workflow.\n\n\n## How to connect your AI agent to GitLab through MCP\n\n\nThe most direct way to supercharge your AI workflow is to give your AI agent native access to `glab` through Model Context Protocol ([MCP](https://about.gitlab.com/topics/ai/model-context-protocol/)).\n\n\n MCP is an open standard that lets AI tools discover and use external capabilities at runtime. Once connected, your AI assistant can read issues, comment on merge requests, check pipeline status, and write back to GitLab, all without copying anything from the UI or writing a single API call yourself.\n\n\n To get started, run:\n\n\n ```shell\n # Start the glab MCP server\n glab mcp serve\n ```\n\n\n Once your MCP client is configured, your AI can answer questions like *\"What's the status of my open MRs?\"* or *\"Are there any failing pipelines on main?\"* by querying GitLab directly, not scraping the web UI, not relying on stale training data. See the [full setup docs](https://docs.gitlab.com/cli/) for configuration steps for Claude Code, Cursor, and other editors.\n\n\n One detail worth knowing: `glab` automatically adds `--output json` when invoked through MCP, for any command that supports it. Your agent gets clean, structured data without you needing to think about output formats. And because `glab` uses the official MCP SDK, it stays compatible as the\n protocol evolves.\n\n\n We've also been deliberate about *which* commands are exposed through MCP. Commands that require interactive terminal input are intentionally\n excluded, so your agent never gets stuck waiting for input that will never come. What's exposed is what actually works reliably in an agent context.\n\n\n ## Let your AI participate in code review\n\n\n Most developers have a backlog of MRs waiting for review. It's one of the most time-consuming parts of the job and one of the best places to put\n AI to work. With `glab`, your agent doesn't just observe your review queue, it can work through it with you.\n\n\n ### See exactly what still needs addressing\n\n\n Start with this:\n\n\n ```shell\n glab mr view 2677 --comments --unresolved --output json\n ```\n\n\n This input returns the full MR: metadata, description, and every\n unresolved discussion, as a single structured JSON payload. Hand that to\n your AI and it has everything it needs: which threads are open, what the\n reviewer asked for, and in what context. No tab-switching, no copy-pasting\n individual comments.\n\n\n \n ```json\n {\n \"id\": 2677,\n \"title\": \"feat: add OAuth2 support\",\n \"state\": \"opened\",\n \"author\": { \"username\": \"jdwick\" },\n \"labels\": [\"backend\", \"needs-review\"],\n \"blocking_discussions_resolved\": false,\n \"discussions\": [\n {\n \"id\": \"3107030349\",\n \"resolved\": false,\n \"notes\": [\n {\n \"author\": { \"username\": \"dmurphy\" },\n \"body\": \"This error handling will swallow panics β€” consider wrapping with recover()\",\n \"created_at\": \"2026-03-14T09:23:11.000Z\"\n }\n ]\n },\n {\n \"id\": \"3107030412\",\n \"resolved\": false,\n \"notes\": [\n {\n \"author\": { \"username\": \"sreeves\" },\n \"body\": \"Token refresh logic needs a test for the expired token case\",\n \"created_at\": \"2026-03-14T10:05:44.000Z\"\n }\n ]\n }\n ]\n }\n ```\n\n\n Instead of reading through every thread yourself, you ask your agent *\"what do I still need to fix in MR 2677?\"* and get back a prioritized summary with suggested changes. This all happens from a single command.\n\n\n ### Close the loop programmatically\n\n\n Once your AI has helped you address the feedback, it can resolve\n discussions:\n\n\n ```shell\n # List all discussions β€” structured, ready for the agent to process\n glab mr note list 456 --output json\n\n # Resolve a discussion once the feedback is addressed\n glab mr note resolve 456 3107030349\n\n # Reopen if something needs another look\n glab mr note reopen 456 3107030349\n ```\n\n\n\n ```json\n [\n {\n \"id\": 3107030349,\n \"body\": \"This error handling will swallow panics β€” consider wrapping with recover()\",\n \"author\": { \"username\": \"dmurphy\" },\n \"resolved\": false,\n \"resolvable\": true\n },\n {\n \"id\": 3107030412,\n \"body\": \"Token refresh logic needs a test for the expired token case\",\n \"author\": { \"username\": \"sreeves\" },\n \"resolved\": false,\n \"resolvable\": true\n }\n ]\n ```\n\n\n\n Note IDs are visible directly in the GitLab UI and API, no extra lookup needed. Your agent can work through the full list, verify each fix, and\n resolve as it goes.\n\n\n ## Talk to your AI about your code more effectively\n\n\n Even if you're not running an MCP server, there's a simpler shift that makes a huge difference: using `glab` to feed your AI better information.\n\n\n Think about the last time you asked an AI assistant to help triage issues or debug a failing pipeline. You probably copied some text from the GitLab UI and pasted it into the chat. Here's what your agent is actually\n working with when you do that:\n\n\n ```text\n open issues: 12 β€’ milestone: 17.10 β€’ label: bug, needs-triage ...\n ```\n\n\n Compare that to what it gets with `glab`:\n\n\n \n ```json\n [\n {\n \"iid\": 902,\n \"title\": \"Pipeline fails on merge to main\",\n \"labels\": [\"bug\", \"needs-triage\"],\n \"milestone\": { \"title\": \"17.10\" },\n \"assignees\": []\n },\n ...\n ]\n ```\n\n\n Structured, typed, complete; no ambiguity, no parsing guesswork. That's the difference between an agent that can act and one that has to ask\n follow-up questions.\n\n\n If you're using the MCP server, you get this automatically: `glab` adds `--output json` for any command that supports it. If you're working directly\n from the terminal, just add the flag yourself:\n\n\n ```shell\n # Pull open issues for triage\n glab issue list --label \"needs-triage\" --output json\n\n # Check pipeline status\n glab ci status --output json\n\n # Get full MR details\n glab mr view 456 --output json\n ```\n\n\n We've significantly expanded JSON output support in recent releases. It now covers CI status, milestones, labels, releases, schedules, cluster agents, work items, MR approvers, repo contributors, and more. If `glab` can\n retrieve it, your AI can consume it cleanly.\n\n\n ### A real workflow\n\n\n ```shell\n $ glab issue list --label \"needs-triage\" --milestone \"17.10\"\n --output json\n ```\n\n\n ```text\n Agent: I found 2 unassigned bugs in the 17.10 milestone that need triage:\n 1. #902 β€” Pipeline fails on merge to main (opened 5 days ago)\n 2. #903 β€” Auth token not refreshing on expiry (opened 4 days ago)\n Both are unassigned. Want me to draft triage notes and suggest assignees based on recent commit history?\n ```\n\n\n ## Your agent is never limited to built-in commands\n\n\n `glab`'s first-class commands cover the most common workflows, but your agent is never limited to them. Through `glab api`, it has authenticated access to the full GitLab REST and GraphQL API surface, using the same session, with no extra credentials or configuration required.\n\n\n This is a meaningful differentiator. Most CLI tools stop at what their commands expose. With `glab`, if GitLab's API supports it, your agent can do it. It's always working from a trusted, authenticated context.\n\n\n A practical example: fetching just the list of changed files in an MR before deciding which diffs to pull in full:\n\n\n ```shell\n # Get changed file paths β€” lightweight, no diff content yet\n glab api \"/projects/$CI_PROJECT_ID/merge_requests/$CI_MERGE_REQUEST_IID/diffs?per_page=100\" \\\n | jq '.[].new_path'\n\n# Then fetch only the specific file your agent needs\nglab api \"/projects/$CI_PROJECT_ID/merge_requests/$CI_MERGE_REQUEST_IID/diffs?per_page=100\" \\\n| jq '.[] | select(.new_path == \"path/to/file.go\")'\n ```\n\n\n ```text\n \"internal/auth/token.go\"\n \"internal/auth/token_test.go\"\n \"internal/oauth/refresh.go\"\n ```\n\n\n For anything the REST API doesn't cover (epics, certain work item queries, complex cross-project data), `glab api graphql` gives you the full\n GraphQL interface:\n\n\n ```shell\n glab api graphql -f query='\n {\n project(fullPath: \"gitlab-org/gitlab\") {\n mergeRequest(iid: \"12345\") {\n title\n reviewers { nodes { username } }\n }\n }\n }'\n ```\n\n ```json\n{\n \"data\": {\n \"project\": {\n \"mergeRequest\": {\n \"title\": \"feat: add OAuth2 support\",\n \"reviewers\": {\n \"nodes\": [\n { \"username\": \"dmurphy\" },\n { \"username\": \"sreeves\" }\n ]\n }\n }\n }\n }\n }\n\n ```\n\n\n Your agent has a single, authenticated entry point to everything GitLab exposes without the token juggling, separate API clients, or configuration\n overhead.\n\n\n ## What's coming and your feedback\n\n\n Two improvements we're actively working on will make `glab` even more useful for agent workflows:\n\n\n **Agent-aware help text.** Today, `--help` output is written for humansvat a terminal. We're updating it to surface the non-interactive alternative\n for every interactive command, flag which commands support `--output json`, and generally make help a useful resource for agents discovering\n capabilities at runtime β€” not just humans.\n\n\n **Better machine-readable errors.** When something goes wrong today, agents get the same human-readable error messages as terminal users. We're\n changing that so errors in JSON mode return structured output, giving your agent the information it needs to handle failures gracefully, retry intelligently, or surface the right context back to you.\n\n\n Both of these are in active development. If you're already using `glab` with an AI tool, you're exactly the audience we want feedback from.\n\n\n * **What friction are you hitting?** Commands that don't behave well in agent contexts, error messages that aren't actionable, gaps in JSON output\n coverage. We want to know.\n\n * **What workflows have you unlocked?** Real usage patterns help us prioritize what to build next.\n\n\n Join the discussion in [our feedback issue](https://gitlab.com/gitlab-org/cli/-/issues/8177) β€” that's where we're shaping the roadmap for agent-friendliness, and where your input will have the most direct impact. If you've found a specific gap, [open an issue](https://gitlab.com/gitlab-org/cli/-/issues/new). If you've got a fix in mind, contributions are welcome. Visit [CONTRIBUTING.md](https://gitlab.com/gitlab-org/cli/-/blob/main/CONTRIBUTING.md) to get started.\n\n\n The GitLab CLI has always been about giving developers more control over their workflow. As AI becomes a bigger part of how we all work, that means making `glab` the best possible interface between your AI tools and your GitLab projects. We're just getting started and we'd love to build the next part with you.\n",[18,19,21],{"featured":29,"template":15,"slug":740},"give-your-ai-agent-direct-structured-gitlab-access-with-glab-cli",{"content":742,"config":750},{"title":743,"description":744,"authors":745,"heroImage":735,"date":747,"body":748,"category":11,"tags":749},"GitHub Copilot's new policy for AI training is a governance wake-up call","Learn what GitHub's Copilot policy change means for regulated industries, and why GitLab's commitment to customer data privacy matters.",[746],"Allie Holland","2026-04-20","GitHub recently [announced](https://github.blog/news-insights/company-news/updates-to-github-copilot-interaction-data-usage-policy/) a significant change to how it handles data from Copilot users. Starting April 24, 2026, interaction data from Copilot Free, Pro, and Pro+ users, including inputs, outputs, code snippets, and associated context, will be used to train AI models by default, unless users actively opt out. Copilot Business and Enterprise customers are exempt under existing contract terms.\n\nFor organizations in regulated industries, including finance, healthcare, defense, and public sector, the policy shift raises questions that go beyond individual developer preferences. It forces a harder look at a question that engineering and security leaders should be asking every AI vendor in their stack: Do you train on our code? \n\nGitLab's answer is no. GitLab does not train AI models on customer code at any tier, and AI vendors are contractually prohibited from using customer inputs or outputs for their own purposes. The [GitLab AI Transparency Center](https://about.gitlab.com/ai-transparency-center/) makes that commitment auditable: a single location documenting which models power which features, how data is handled, subprocessor relationships, and data retention periods. The GitLab AI Transparency Center also lists the compliance status of each feature, including confirmation that GitLab's current AI features do not qualify as high-risk systems under the EU AI Act. It's a standard GitLab CEO Bill Staples has consistently [reiterated](https://www.linkedin.com/posts/williamstaples_gitlab-1810-agentic-ai-now-open-to-even-activity-7443280763715985408-aHxf?utm_source=share&utm_medium=member_desktop&rcm=ACoAABsu7EUBcb_a1-JHKS9RC0B5rf8Ye-5XM60) and one reflected in GitLab's mission and [Trust Center](https://trust.gitlab.com/).\n\n## What the policy change actually means\n\nGitHub's announcement also specifies that the data may be shared with GitHub affiliates, including Microsoft, for AI development purposes.\n\nA policy change of this nature forces organizations to re-examine their AI governance posture, audit their Copilot license tiers, and confirm that the right controls are configured across their teams.\n\n## Why AI governance matters in regulated environments\n\nSource code is often among an organization's most sensitive intellectual property. It may contain references to internal systems, reflect proprietary business logic, or touch data flows governed by strict retention and access policies. When that code passes through an AI assistant, questions about training data usage, model vendor relationships, and data residency become compliance concerns.\n\nThe exposure is particularly acute for financial services firms that have invested in proprietary algorithms, fraud detection logic, credit risk models, underwriting rules, trading strategies. When AI tooling processes that code and uses it to train models serving competitors, vendor data practices become an IP concern.\n\nFinancial institutions operating under [the Federal Reserve's Supervisory Guidance on Model Risk Management (SR 11-7) and the](https://www.federalreserve.gov/supervisionreg/srletters/sr1107.htm) [Digital Operational Resilience Act (DORA)](https://eur-lex.europa.eu/eli/reg/2022/2554/oj/eng) are required to maintain documented, auditable oversight of third-party technology providers, including understanding how those providers handle data. Third-party AI tools used in development workflows increasingly fall within the scope of model risk oversight, and material changes to vendor data practices require updated documentation. \n\nIn the public sector, [the National Institute of Standards and Technology Special Publication 800-53 (NIST 800-53)](https://csrc.nist.gov/publications/detail/sp/800-53/rev-5/final) and the [Federal Information Security Modernization Act (FISMA)](https://www.cisa.gov/topics/cyber-threats-and-advisories/federal-information-security-modernization-act) establish that sensitive or classified code must never leave a controlled boundary. For U.S. Department of Defense and intelligence community environments in particular, a vendor's default data posture is an operational concern. In healthcare, [the Health Insurance Portability and Accountability Act (HIPAA)](https://www.hhs.gov/hipaa/index.html) governs how patient-adjacent data is handled by third parties, and development environments that touch clinical systems increasingly fall within that scope.\n\nAcross all of these contexts, the common thread is the same: A vendor policy that changes data usage defaults, requires individual opt-out, and offers different protections depending on account tier introduces exactly the kind of uncontrolled variable that compliance teams cannot afford.\n\n## What regulated industries actually need from AI vendors\n\nRegulated organizations have largely moved past debating whether to adopt AI in development workflows. The focus now is on doing so in a way they can defend to regulators, boards, and customers. That shift has surfaced a consistent set of requirements regardless of sector.\n\n**Contractual certainty.** Regulated firms need to know, with specificity, what happens to their data. A clear, documented, unconditional commitment is what's required, not something that varies by plan or requires action before a deadline.\n\n**Auditability.** Model risk management frameworks require organizations to understand and validate the AI systems they deploy, including the training data behind those models and the third parties involved in their development. Vendors who cannot answer these questions create documentation risk for the organizations relying on them.\n\n**Separation from vendor incentives.** When an AI vendor trains models on customer usage data, code and workflows become inputs to a system that also serves competitors. For institutions with proprietary trading logic, underwriting models, or fraud detection systems, that's a genuine IP exposure.\n\n## GitLab's position on AI data governance\n\nGitLab does not use customer code to train AI models. This commitment applies at every tier, and AI vendors are contractually prohibited from using inputs or outputs associated with GitLab customers for their own purposes.\n\nThis is a deliberate architectural and policy choice, not a feature of a particular pricing tier. As GitLab's [post on enterprise independence](https://about.gitlab.com/blog/why-enterprise-independence-matters-more-than-ever-in-devsecops/) notes, data governance has become \"an increasingly critical factor in enterprise technology decisions, driven by a complex web of national and regional data protection laws and growing concern about control over sensitive intellectual property.\"\n\nGitLab is also cloud-neutral and model-neutral while supporting self-hosted deployments, not commercially tied to any single cloud provider or large language model (LLM). That i[ndependence matters](https://about.gitlab.com/blog/why-enterprise-independence-matters-more-than-ever-in-devsecops/) for regulated organizations evaluating vendor concentration risk. The [AI Continuity Plan](https://handbook.gitlab.com/handbook/product/ai/continuity-plan/) documents how vendor changes are managed, including material changes to how AI vendors treat customer data, a direct response to the governance requirements under frameworks like [DORA](https://handbook.gitlab.com/handbook/legal/dora/). \n\n## The governance gap AI teams need to close\n\nGitHub's policy update is a reminder that for organizations in regulated industries, understanding exactly how an AI tool handles data is a prerequisite for using it at all. That means asking vendors for clear, documented answers: Is our data used for model training? Who are your AI model subprocessors? What happens if a vendor changes its data practices? Can we deploy in a way that keeps all AI processing within our own infrastructure? What indemnification do you offer for AI-generated output?\n\nVendors who can answer those questions clearly, and document those answers in an auditable form, are vendors you can build on. **Those who cannot will create compliance debt every time they ship a policy update.** And when a vendor can change its data practices with 30 days notice, that's not a partnership built for regulated industries. That's a liability.\n\n> Learn more about GitLab's approach to AI governance at the [GitLab AI Transparency Center](https://about.gitlab.com/ai-transparency-center/).",[18,19],{"featured":14,"template":15,"slug":751},"github-copilots-new-policy-for-ai-training-is-a-governance-wake-up-call",{"promotions":753},[754,767,778,790],{"id":755,"categories":756,"header":757,"text":758,"button":759,"image":764},"ai-modernization",[11],"Is AI achieving its promise at scale?","Quiz will take 5 minutes or less",{"text":760,"config":761},"Get your AI maturity score",{"href":762,"dataGaName":763,"dataGaLocation":243},"/assessments/ai-modernization-assessment/","modernization assessment",{"config":765},{"src":766},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1772138786/qix0m7kwnd8x2fh1zq49.png",{"id":768,"categories":769,"header":770,"text":758,"button":771,"image":775},"devops-modernization",[19,570],"Are you just managing tools or shipping innovation?",{"text":772,"config":773},"Get your DevOps maturity score",{"href":774,"dataGaName":763,"dataGaLocation":243},"/assessments/devops-modernization-assessment/",{"config":776},{"src":777},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1772138785/eg818fmakweyuznttgid.png",{"id":779,"categories":780,"header":782,"text":758,"button":783,"image":787},"security-modernization",[781],"security","Are you trading speed for security?",{"text":784,"config":785},"Get your security maturity score",{"href":786,"dataGaName":763,"dataGaLocation":243},"/assessments/security-modernization-assessment/",{"config":788},{"src":789},"https://res.cloudinary.com/about-gitlab-com/image/upload/v1772138786/p4pbqd9nnjejg5ds6mdk.png",{"id":791,"paths":792,"header":795,"text":796,"button":797,"image":802},"github-azure-migration",[793,794],"migration-from-azure-devops-to-gitlab","integrating-azure-devops-scm-and-gitlab","Is your team ready for GitHub's Azure move?","GitHub is already rebuilding around Azure. Find out what it means for you.",{"text":798,"config":799},"See how GitLab compares to GitHub",{"href":800,"dataGaName":801,"dataGaLocation":243},"/compare/gitlab-vs-github/github-azure-migration/","github azure migration",{"config":803},{"src":777},{"header":805,"blurb":806,"button":807,"secondaryButton":812},"Start building faster today","See what your team can do with the intelligent orchestration platform for DevSecOps.\n",{"text":808,"config":809},"Get your free trial",{"href":810,"dataGaName":50,"dataGaLocation":811},"https://gitlab.com/-/trial_registrations/new?glm_content=default-saas-trial&glm_source=about.gitlab.com/","feature",{"text":506,"config":813},{"href":54,"dataGaName":55,"dataGaLocation":811},1777493585702]