Glossary

Model Context Protocol (MCP)

The Model Context Protocol (MCP) is an open protocol for connecting AI agents to tools and data sources. A tool server lists the tools it offers, and an agent calls them with structured inputs.

In a run

In the supplier follow-up run, the ERP, the supplier mailbox, the document store and accounting are each a tool the agent reaches through Ares. Each call is checked against the access policy, and a tool or action outside the credential for the job is refused.

Our word for it

Governed communication: a controlled path between agents and tools.

How MCP fits a workflow

In a workflow, the agent decides the next step and the tools do the work in your systems. MCP is the path between the two, so it is where each action can be checked. Ares sits on that path: the call, the decision and the reason all go on the record.

Tools are permissions

A tool server changes when someone deploys it: tools are added, renamed and removed. Each new tool is a new permission. Ares keeps the published tool card as the ceiling, so a change on the server never widens what an agent may do.

Free tools for MCP servers

Where it stops

Ares controls what the agent may call. What runs inside the tool server is outside Ares, and stays the job of the server and its runtime.

Watch an agent work across four tools in the sandbox.

Related terms

  • Tool callingTool calling is how an AI model acts outside its own text: it picks a tool, fills in the inputs and asks for the call.
  • Tool cardOn Ares, a tool card is the published description of a tool server: its address, the tools it offers and the inputs each tool takes.

Tell us about one workflow.

We will say whether an agent should do it, and what the first step costs.

A senior member of the team replies within two business days.