Applaud supports the Model Context Protocol (MCP), a standard way for AI applications to connect to systems and use the capabilities they provide.
MCP gives AI applications a consistent way to discover and use tools across different systems. In Applaud, you can use MCP in two directions:
- Inbound MCP: External AI applications connect to Applaud through Applaud MCP Server.
- Outbound MCP: Applaud connects to an external MCP server through the MCP Server connector.
The direction you need depends on where the AI application is running and which system provides the capabilities you want to use.
Choose the right MCP option
| If you want to... | Use |
| Let an external AI application connect to Applaud and use supported Applaud capabilities | Applaud MCP Server |
| Connect Applaud to an external MCP server and use its tools with Applaud AI experiences | MCP Server connector |
| Use MCP as a source for employee records in Employee Feeds | Employee Feeds MCP |
Key MCP concepts
Understanding these terms will help you work with MCP in Applaud.
MCP server
An MCP server provides capabilities that AI applications can discover and use.
The server can represent a system such as an HR platform, business application, or service that provides tools for AI applications.
MCP client
An MCP client is the AI application or agent that connects to an MCP server to use its capabilities.
For inbound MCP, an external AI application acts as the MCP client and connects to Applaud.
For outbound MCP, Applaud connects to an external MCP server and uses the tools it provides.
MCP tool
An MCP tool is an individual capability provided by an MCP server. An AI application can use the tool to perform a supported action or retrieve information.
For example, an external MCP server might provide tools for creating a document, searching a business system, or updating a record.
Use inbound MCP to connect external AI applications to Applaud
With Applaud MCP Server, Applaud provides an MCP endpoint that external AI applications can connect to.
The external application can then discover and use the Applaud capabilities available through MCP.
The flow looks like this:
External AI application → Applaud MCP Server → Applaud capabilities
This is useful when you want an AI application outside Applaud to work with supported Applaud capabilities through a standard MCP connection.
Access to Applaud capabilities continues to follow the permissions and controls that apply to the connected user and the capabilities being used.
Use outbound MCP to connect Applaud to external systems
With the MCP Server connector, Applaud connects to an external MCP server and discovers the tools it provides.
You can then configure the tools you want to use in Applaud and make them available to supported AI experiences, such as AI agents and agentic workflows.
The flow looks like this:
Applaud → MCP Server connector → External MCP server → MCP tools
A typical setup is:
- Connect to the external MCP server.
- Discover the tools it provides.
- Configure the tools you want to use.
- Test the tools.
- Activate them.
- Attach them to the relevant AI experience.
This gives you a way to extend what Applaud AI experiences can do by connecting them to capabilities provided by other systems.
MCP and Agent Tools
Agent Tools is where you manage tools that Applaud AI experiences can use.
When you connect an external MCP server through the MCP Server connector, its available tools can be discovered in Agent Tools. You can then configure and manage the tools you want Applaud AI experiences to use.
Authentication and access
Authentication depends on the MCP connection and the requirements of the server or client involved.
For an outbound connection, the MCP Server connector supports the authentication options available for the external MCP server.
For an inbound connection, access to the Applaud MCP Server is controlled by how Applaud and the external AI application are configured.
Always ensure that the authentication method, permissions, and access controls align with your organization's security requirements.
Employee Feeds and MCP
MCP can also be used with Employee Feeds to provide employee data from an MCP-enabled source.
This is a separate use case from connecting MCP tools to AI experiences. Employee Feeds uses MCP to retrieve employee records as part of the employee data feed process.