Agent Ready

MCP

Model Context Protocol: the open standard most connectors use to offer their tools to an AI assistant.

MCP (Model Context Protocol) is an open standard for connecting AI assistants to tools and data. Anthropic introduced it on November 25, 2024, and ChatGPT, Claude, Cursor, Grok Bot and Muse all support it today.

An MCP server describes what it can do as a list of tools. Each tool has a name, a description, an input schema and optional hints such as read-only or destructive. The assistant reads those descriptions to decide which tool to call, which is why the words in a tool description matter so much for whether a connector gets used.

Because the protocol is shared, one MCP server can power a connector in several stores. Each store still has its own listing, review and rules.

Claude Code: what is loaded in a chatClaude Code's context panel: MCP tools 20.6k tokens, skills 9.9k tokens, and 356.6k tokens of MCP tools deferred

Always loaded

Tool and skill names and descriptions. They sit in the AI's context at the start of every chat.

Deferred

Full MCP tools and skills. Their details load only when a request needs them.

Tool names and descriptions sit in the AI’s context in every chat. That’s why they decide whether a connector gets used.

Sources