Enrich records with an internal tool
An in-house MCP server that knows council districts or service zones is called for each 311 request before it is routed.
The Model Context Protocol (MCP) is an open standard that lets software expose tools for AI assistants and other programs to call. A growing number of software vendors publish MCP servers alongside their REST APIs.
SentraBridge can call the tools of any streamable HTTP MCP server inside a flow: look something up for each record, or hand records to another system's tool. Servers that need a person to sign in with OAuth are supported. Records pass through; nothing is stored.
| resource | read | write | changes only | fields |
|---|---|---|---|---|
| Custom paths | any GET path | POST, PUT, PATCH, DELETE | with a changed-since parameter | Fields from the response |
An in-house MCP server that knows council districts or service zones is called for each 311 request before it is routed.
Approved records are sent to a vendor system that only offers an MCP server, not a REST API.
Tools the IT team built for staff AI assistants are reused in scheduled flows, with the same sign-in and permissions.
A classification tool on an MCP server suggests a category for free-text requests, checked by rules before writing.
Systems cities commonly connect with MCP server, each with its own connector page.
A server that exposes tools over the Model Context Protocol, an open standard for connecting software to AI assistants and other clients. Each tool has a name, inputs and a result.
Yes. A flow calls a tool directly with mapped arguments, the same way it calls a REST endpoint. No model is involved unless the tool itself uses one.
Streamable HTTP servers, signed in with a Bearer token, an API key header, OAuth client credentials, or OAuth where a person signs in.
We'll set up a workspace with your systems in a sandbox and build your first flow with you.