AI Library: New MCP Tools

Enabling a secure, governed AI ecosystem is essential as organizations integrate AI assistants and autonomous agents into their daily operations. To help you safely authorize agents to do real work, we are excited to announce a host of new MCP toolkits, accessible right from the AI Library.  

Whether you’re using external models or custom agents built in Domo, you can now manage your workflows, alerts, and workspaces using simple, conversational commands. Agents operate strictly under your existing platform permissions, ensuring they only access what you’re already authorized to see.  

Let’s explore what’s new!

MCP Toolkits: Workflows & Automation

Building and maintaining business automations across multiple products can require significant administrative effort, technical domain knowledge, and constant context-switching. When valuable features are disconnected from AI, organizations struggle to leverage models to streamline processes.  

That changes with our new MCP automation tools, which equip AI with secure, standardized capabilities to design, manage, and run your workflows, Code Engine packages, and Task queues.  

These tools let you:  

• Accelerate automation delivery: Enable AI to design, update, and launch production-ready automations instantly using natural language commands.  

• Empower business users: Lower the barrier to entry and allow non-technical users to build process automations without deep development expertise.  

• Increase operational efficiency: Reduce administrative work, manual maintenance, and context-switching across tools.  

• Improve governance and consistency: Ensure AI safely leverages standardized automation tools and configurations across your entire platform.  

• Maximize platform ROI: Orchestrate and connect existing business capabilities so you can get more out of your automation investments.  

Available toolkits:  

Workflow Toolkit (17 tools): Manage the complete workflow lifecycle, allowing AI to create, validate, test-run, and deploy automated definitions.  

Code Engine Toolkit (6 tools): Enable AI to package custom business logic, version unreleased code, and securely execute background functions.  

Task Center Queue Toolkit (10 tools): Coordinate human-in-the-loop approvals, letting AI manage queues, review tasks, and monitor watchlists.

Forms Toolkit (3 tools): Support structured data collection by allowing AI to build, retrieve, and manage interactive workflow forms.  

Workflow Design Skills (17 skills): Teach AI to design error-free workflows with guidance on conditional gateways, timers, and data layouts.

MCP Toolkits: Alerts

Alerts tend to get created after something has already gone wrong, but it can be tedious to manually set up notifications for metrics to watch.  

With our new MCP Toolkits for Alerts, you can ask an agent to monitor certain metrics, and it will handle the creation and management of Alerts and subscriptions, including the threshold and who gets notified. Even better, the alert will run on Domo's own detection rather than on the agent. Its condition is evaluated as your data refreshes, so nothing calls a model or spends tokens while the alert waits.

MCP Toolkits: Workspaces

Organizing content and coordinating team access is essential but often demands continuous administrative effort and manual upkeep.  

Our new MCP Toolkits for Workspaces allow agents to securely create and manage these spaces on your behalf. They can now handle the creation, retrieval, and updating of Workspaces and folders. Every automated action strictly respects your platform permissions on underlying resources, ensuring you can safely offload administrative tasks without compromising security.

MCP Toolkits: Sandbox

Safely promoting dashboards, apps, and dataflows to production is essential, but manually managing the repository workflows to get them there can be time-consuming.

Our new Sandbox MCP Toolkit introduces dedicated tools to create, commit, promote, and delete repositories, as well as manage sharing of them. To coordinate these actions, a new universal orchestration skill serves as your conversational entry point, routing requests to the appropriate tool based on your intent. It interactively gathers configuration options upfront (such as version selections) and prompts you for any missing inputs, guiding you through the correct configurations when an action is invoked.  

Check out all our new toolkits in the AI Library and learn more about connecting AI tools to Domo using MCP.