Cluster
The Remote MCP Server
A remote MCP server is a hosted endpoint your agent connects to over the Model Context Protocol, calling its tools directly with no SDK to wire and no infrastructure to run. This cluster covers what remote and hosted MCP mean, how an agent invokes a tool, and why enrichment ships over MCP instead of a REST API you wrap yourself.
5 questions about the remote mcp server
- How does an agent call a remote MCP tool?
An agent calls a remote MCP tool by connecting to the HTTP endpoint, running the initialize handshake, then sending a tools/call request with arguments.
- Remote MCP vs a REST API: which is better for agents?
For agents, a remote MCP server beats a bare REST API because the protocol carries tool schemas the model reads and calls with no per-host adapter.
- What is a hosted MCP server?
A hosted MCP server is one the vendor runs on its own infrastructure, reachable over the network by URL, so your agent connects without installing anything.
- What is a remote MCP server?
A remote MCP server is a Model Context Protocol server hosted at a URL that any agent reaches over the network, not spawned locally over stdio.
- Why expose enrichment over MCP instead of an API?
Agents speak MCP natively. Exposing enrichment as an MCP tool drops the client code, auth, and schema wiring an agent needs to consume a REST API.
Related clusters
Frequently asked questions
Remote MCP vs a REST API: which is better for an agent?
For an agent, MCP. A REST API makes you write the client, parse the schema, and handle auth on every call. An MCP tool arrives typed and self-describing, so the agent invokes it the moment it connects.
Why expose enrichment over MCP instead of an API?
Because the agent is the consumer. Enrichment over MCP lands as a callable tool inside the agent's loop, with no glue code between the request and the dossier. An API leaves that integration for you to build and keep alive.
Sniff a domain.
Run sniff_domain on any site and read its five-layer dossier in one call.