Model Context Protocol
APIs and agents, a common language.
The Model Context Protocol lets AI applications discover and use external tools. Let’s talk about the data and APIs your agent needs.
How the protocol works
Discover. Query. Interpret.
Discover tools
An MCP client reads the catalog of a compatible server. Each tool describes its purpose and the parameters it accepts.
Send parameters
The client requests a tool using structured data. The integration defines the required authentication and permissions.
Receive a result
The tool returns content that the agent can use in its workflow. The format and error handling are part of the contract.
MCP and REST
Choose an interface for your workflow.
MCP for agents
- The assistant discovers tools and their parameters.
- A compatible client manages communication with the server.
- The integration defines which tools the agent can use.
REST for applications
- Your code selects the endpoint and parameters for each request.
- Your backend controls the sequence and error handling.
- Each API’s documentation defines its contract.
About your integration
What is MCP?
The Model Context Protocol is an open protocol for AI applications to exchange information with tools and data sources. A client and server must agree on the capabilities they use.
What does a remote client need?
Check which transport and authentication methods your client supports. An integration based on Streamable HTTP and OAuth requires support for both and a configuration compatible with the server.
What can we discuss?
Tell us which AI client you use, what data you need and how your process works. With that context, we can review the requirements of an MCP integration.
Where can I find APIs and country coverage?
The API Catalog describes the products and their coverage. Include the APIs and countries you are interested in when you contact us.