MCP & INTEGRATIONS
Connect your local AI system to the tools, applications, and data it needs to perform useful work. Learn how MCP and other integrations turn individual models into flexible, connected workflows.
WHAT MCP DOES
The Model Context Protocol, or MCP, provides a standard way for AI applications to connect with external tools and information. Instead of building a separate custom connection for every model and application, MCP allows compatible clients and servers to work together through a shared protocol.
An MCP client runs inside an AI application, while an MCP server exposes specific tools, resources, or reusable prompts. The AI application can then discover those capabilities and use them when they are relevant to your request.
MOVE BEYOND A STANDALONE CHATBOT
A model can generate useful responses on its own, but integrations allow it to interact with files, software, services, and live information. This makes it possible to build practical workflows for research, content creation, coding, automation, Blender, Unreal Engine, and other specialized tasks.
HOW MCP IS STRUCTURED
HOSTS, CLIENTS, AND SERVERS
An MCP host is the AI application you interact with. Inside it, an MCP client manages the connection to one or more MCP servers. Each server provides a focused set of capabilities, such as reading files, searching a knowledge base, controlling software, or accessing an external service.
CORE MCP CAPABILITIES
TOOLS, RESOURCES, AND PROMPTS
MCP servers can expose tools that perform actions, resources that provide information, and prompts that package reusable instructions or workflows. The available capabilities depend on the server, and the host controls how they are presented to the model and the user.
CONNECTION TYPES
LOCAL AND REMOTE MCP SERVERS
Local MCP servers run on your own computer and can connect AI applications to local files, software, or hardware. Remote servers run elsewhere and provide access through a network connection. Local servers offer greater privacy and control, while remote servers can make cloud services and shared systems easier to access.
BEYOND MCP
OTHER WAYS TO CONNECT YOUR STUDIO
Not every integration uses MCP. APIs, plug-ins, command-line tools, webhooks, shared folders, and custom scripts can also connect models with applications and services. The right method depends on the software involved, the actions you want to perform, and whether the workflow must remain entirely local.
PLAN THE CONNECTION
START WITH THE WORKFLOW
Begin with the task you want to complete rather than installing integrations at random. Identify the model, application, files, and services involved, then determine which connections are actually required. A focused workflow is easier to configure, test, secure, and maintain.
SECURITY & PERMISSIONS
LIMIT ACCESS BY DEFAULT
Every connection expands what an AI application may be able to read or change. Grant access only to the files, folders, tools, and accounts required for the workflow. Review each server’s source, configuration, and permissions before connecting it, and require confirmation before allowing sensitive or destructive actions.
BUILD RELIABLY
CONNECT AND TEST ONE PIECE AT A TIME
Add one server or integration, confirm that it connects correctly, and test its individual capabilities before expanding the workflow. Record the installation steps, settings, permissions, and software versions as you go. This makes problems easier to isolate and gives you a repeatable setup you can restore or move to another system.
CREATIVE WORKFLOWS
CONNECT AI WITH BLENDER, UNREAL, AND MORE
Creative integrations can help generate concepts, organize references, write scripts, manage project files, and automate repetitive production tasks. Keep the artist in control of important visual and technical decisions, and test generated actions on copies or version-controlled projects before using them in production.
YOUR FIRST INTEGRATION
BEGIN WITH A LOW-RISK LOCAL WORKFLOW
Start by connecting an AI application to a dedicated test folder containing non-sensitive files. Practice listing, reading, searching, and creating files within that limited space. Once the behavior is predictable, you can gradually introduce additional tools and carefully expand access.
MAINTAIN YOUR CONNECTIONS
DOCUMENT VERSIONS, SETTINGS, AND ACCESS
Keep a simple record of every integration you install, including its source, version, purpose, configuration, permissions, and update procedure. Review the list periodically, remove connections you no longer use, and retest important workflows after updating models, applications, or servers.
FOUNDATION COMPLETE
BUILD A WORKING AI STUDIO
With your base system, hardware, software stack, models, and integrations understood, you are ready to combine them into complete workflows. Continue to Studios to explore practical setups for image generation, video, Blender, Unreal Engine, automation, and more.