Glossary
What is Model Context Protocol (MCP)?
Model Context Protocol (MCP) is an open standard that lets AI assistants connect to tools and data in a consistent way.
What it is
MCP defines how an AI application asks a separate server for tools. Those tools might search a database, read a file or create a task. Any tool server that speaks the standard can work with any assistant that supports it. Tool builders write one server, not one connector per assistant.
Why it matters for businesses
You can give an assistant access to your systems without rebuilding each connection. You also decide which tools are exposed. Keep the list short, and review what each tool is allowed to change.
Risks to check
A tool server can read data you did not intend to share, or take actions you did not approve. Connect only servers you trust, and check each tool’s permissions. Require approval for any tool that changes data.
Example
A Riyadh accounting firm connects an MCP server that offers read-only access to client invoices. Its AI assistant can answer questions about overdue balances. The firm does not let the assistant send reminders without a person approving them first.
Montr’s AI agent can use external MCP servers. An admin must approve each one, and it applies to a single brand.