Model Context Protocol (MCP)
Overview
Model Context Protocol is a standardized interface specification that defines how AI models communicate with external systems to access context, execute actions, and retrieve information. In the Knovvu Virtual Agent, MCP serves as the foundational layer that enables agents to extend their capabilities beyond their training data, accessing real-time information and performing actions in connected systems.
Instead of building custom connectors for each system, MCP provides a universal protocol that any service can implement, creating an ecosystem of interoperable components that agents can leverage dynamically.
Configuration
- On the Integrations / MCP Servers page, create a connection by entering the MCP Server Name, MCP Server URL, and Connection Type. There are two available connection types: Streamable HTTP (recommended) and SSE (legacy connection type). If authentication is required, enable the toggle and fill in the Header Name and Header Value fields.
- On the Agent Details page, click the Import from MCP button.
- After selecting the MCP server, the available MCP tools are automatically fetched and displayed. Select the desired tools using the checkboxes, then click Import to complete the process.
- After importing the selected tools, they appear under the Tools section, with details such as the tool name, description, and parameter information automatically populated.
Work & Data Flow
