liveBook · Manning

MCP

As explained in Claude at Work

The Model Context Protocol is an open standard for connecting AI models to external systems through discoverable, permissioned operations.

MCP connects Claude to external systems through lightweight middleware called an MCP server. The server lists available operations and supports both reading data and making changes. section 3.5

A connector uses several safeguards:

  • Authenticate directly with the connected system; Claude receives an OAuth token that can be revoked there. section 3.5
  • Set each operation to run automatically, ask for confirmation, or stay blocked; changes can require approval. section 3.5

Review a write action

Claude can draft an email response, then request approval before the connector sends or otherwise changes data. section 3.5

Figure 3.7 The MCP server is the adapter between Claude’s tool list and the source system. It holds the credentials and exposes a small set of permissioned actions.
CH03 F07 Stoneman

See how the MCP server sits between Claude’s tool list and the connected system. section 3.5

How Claude at Work covers it

How deep it goes
Describes middleware, operation discovery, read/write access, OAuth sign-in, and per-action controls.
Examples
Uses email search, ticket updates, and a drafted email requiring approval as examples.
What you can do after
Configure connectors in Claude Desktop and choose whether operations run, require confirmation, or remain blocked.
How it connects
Links external-system connectors with OAuth authentication and action-level permissions.

Also explained in

Compare how 21 books explain MCP →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Claude at Work. The section links point to the parts of the book it draws on.