What is an MCP server?
An MCP server is a program that implements the Model Context Protocol to expose tools and data to AI agents over a standard interface, so a model like Claude can read context and take real actions in the loop.
A phone system agents can drive.
VanceTel ships a Model Context Protocol server so AI agents and Claude Code can act on your phone system directly — send a text, resolve a contact, or check a message's status. It sits alongside the REST API and HMAC-signed webhooks, so models, code and people all work through the same surface.
An MCP server in practice.
A developer connects Claude Code to the VanceTel MCP server. They ask it to text a customer a confirmation. The agent calls the send-message tool, the text goes out from the business number, and it also lands in the softphone inbox — all without the developer writing a single request by hand.
Keep reading.
What does an MCP server do?
An MCP server speaks the Model Context Protocol, a standard way for AI agents to discover and call tools. Instead of bespoke glue for every model, you expose your capabilities once over MCP, and any MCP-aware agent can use them — reading data and taking actions through a defined interface.
What can the VanceTel MCP server do?
It lets AI agents and Claude Code act on your phone system in the loop — send a text, resolve a contact, or check a message's status — alongside the REST API and webhooks. The agent works through the same programmable surface your own code uses.
Is an MCP server the same as a REST API?
They overlap but serve different callers. A REST API is for your own software; an MCP server is the agent-facing front door that lets a model invoke tools conversationally. VanceTel ships both, so humans, code and agents drive the same phone system.