The MCP Server connector lets you connect Applaud to an external Model Context Protocol (MCP) server and use the tools it provides in Applaud.
MCP provides a standard way for AI applications to connect to external capabilities. Instead of building a separate integration for every system, an MCP server can make its capabilities available as tools that an AI agent can discover and use.
In Applaud, you can connect to an MCP server, discover its tools, and configure those tools for use by your AI agents, case experiences, and workflows.
What is MCP?
Model Context Protocol (MCP) is a standard that allows AI applications to connect to external systems and use the capabilities they provide.
An MCP server makes a set of tools available. For example, a service might provide tools for searching information, creating records, checking status, or performing other actions.
The MCP Server connector lets Applaud connect to that server and bring its tools into Agent Tools.
Think of it as a bridge between Applaud and an external MCP server:
External MCP server → MCP tools → Applaud Agent Tools → AI agents and workflows
This means you can give an Applaud agent access to capabilities that are already provided by an external system, without building those capabilities from scratch in Applaud.
How the MCP Server connector works
The MCP Server connector follows the same general pattern as other Applaud connectors:
Connect → Discover → Configure → Test → Activate → Attach
Connect
Create an MCP Server connection using the URL of the external MCP server and the authentication method it requires.
Discover
Once connected, Applaud can discover the tools provided by the MCP server.
These tools describe the actions the external system makes available to AI applications.
Configure
Select the tools you want to use in Applaud and configure how they should appear and behave as Applaud tools.
You can add information such as descriptions, risk levels, activity messages, and conditions to help your AI agents use the tools appropriately.
Test and activate
Test the tools to make sure they work as expected, then activate them when they're ready to use.
Attach
Attach active tools to the AI agents, case experiences, or workflows that need to use them.
Connecting an MCP server doesn't automatically make all of its tools available to employees. You choose which tools to bring into Applaud and where they can be used.
When to use the MCP Server connector
Use the MCP Server connector when an external system provides an MCP server, and you want Applaud agents to use its capabilities.
For example, you might use it when:
- A software vendor provides an MCP server that exposes useful actions.
- Your organization has built its own MCP server.
- You need to connect to a service that doesn't have a dedicated Applaud connector.
- An existing MCP server provides capabilities that you want your AI agents to use.
- You want to reuse capabilities that are already exposed through MCP instead of creating a separate integration.
If Applaud already provides a dedicated connector for the system you want to use, that connector may provide a more straightforward experience. Use the MCP Server connector when a generic MCP connection is the better fit.
MCP servers and MCP tools
It's useful to understand the difference between an MCP server and an MCP tool.
MCP server
The external service that provides capabilities through MCP.
MCP tool
An individual capability provided by an MCP server. For example, an MCP server might provide separate tools for searching records and creating records.
Applaud MCP tool
The tool you configure in Agent Tools based on a capability provided by the external MCP server. You can refine how the tool is presented and control where it can be used in Applaud.
Agent Tools
The Applaud area where you manage connections and tools and prepare them for use by AI agents and other experiences.
This separation lets you connect to an MCP server once and choose which of its capabilities you want to make available in Applaud.
How MCP tools work with AI agents
Once an MCP tool is active and attached to an AI agent, the agent can decide when to use it based on the user's request and the instructions you've configured.
For example, a user might ask an AI assistant to perform an action that requires information from an external system. If the appropriate MCP tool is available, the agent can use that tool to interact with the external system and return the result to the user.
The external MCP server remains responsible for performing the underlying action. Applaud provides the connection and makes the capability available to the agent through Agent Tools.
Outbound and inbound MCP
The MCP Server connector provides an outbound connection.
This means:
Applaud → external MCP server
Applaud connects to an MCP server hosted outside Applaud and uses the tools that server provides.
This is different from inbound MCP, where an external AI application connects to Applaud's MCP endpoint to use capabilities provided by Applaud.
For example:
- Outbound MCP: Applaud connects to a vendor's MCP server.
- Inbound MCP: An external AI client connects to Applaud.
These are separate capabilities and serve different purposes.
Before you configure an MCP Server connector
Before creating a connection, make sure you have:
- The URL for the external MCP server.
- The authentication details required by the server.
- The appropriate permissions to configure Agent Tools.
- A clear understanding of which capabilities you want to make available to your AI agents.
Keep in mind that connecting an MCP server is only the first step. You still need to discover the available tools, configure and test the ones you want to use, activate them, and attach them to the appropriate AI agents or workflows.