MCP
The Model Context Protocol is an open standard protocol that defines how AI applications connect with external tools, data sources, and other contextual resources.
Grokking AI Applications
The Model Context Protocol is an open standard for discovering and invoking tools exposed by servers for use by AI applications and agents.
- How it is used
- Deploys interoperable GenAI applications by standardizing how assistants discover and use tools exposed by servers.
- How deep it goes
- Explains servers, clients, discovery, manifests, handshakes, and invocation, then demonstrates implementations with Langflow and Python.
- Examples
- Uses knowledge searches, customer orders, culinary itineraries, recipes, email, and a calculator as discoverable tools.
- What you can do after
- Build it from scratch with FastMCP or apply it through Langflow to expose, discover, test, and invoke reusable tools.
AI Agents in Action, Second Edition
The Model Context Protocol is an open protocol for connecting AI applications and models with external tools, data sources, and contextual resources.
- How it is used
- Connects agents to reusable external tools, data, and services instead of requiring separate integrations in each agent.
- How deep it goes
- Works through discovery, local and remote connections, and converting typed tools into servers, including latency and failure trade-offs.
- Examples
- Uses a FastMCP journal server and a Python agent launching a local research server over STDIO.
- What you can do after
- Build it from scratch: expose typed functions through FastMCP and register an MCP server with an agent.
Build AI-Enhanced Web Apps
A standardized protocol for connecting AI applications and agents with externally provided tools, prompts, resources, and data sources.
- How it is used
- Standardizes how AI applications discover and invoke externally provided capabilities without custom wiring for each integration.
- How deep it goes
- Works through client–server roles and a local stdio implementation, while directing readers elsewhere for the current specification.
- Examples
- Builds a joke-fetching MCP server and connects it to a Next.js chat application through stdio.
- What you can do after
- Build it from scratch: expose an API-backed MCP tool and connect it to a Next.js assistant using the Vercel AI SDK.
Build an AI Agent (From Scratch)
Model Context Protocol is an open protocol that standardizes how AI applications connect to tools, prompts, resources, and external services.
- How it is used
- Standardizes how an agent discovers and invokes external tools instead of maintaining a separate integration for each one.
- How deep it goes
- Explains host–client–server roles and both interfaces, then walks through a FastMCP server and client connection; focuses on tools rather than prompts or resources.
- Examples
- Uses third-party service integrations and a Python web-search server whose discovered tool is checked through an MCP client.
- What you can do after
- Build it from scratch with FastMCP, then connect a client to discover and invoke the server’s tool.
AI Agents and Applications
The Model Context Protocol is a standard protocol for services to expose tools that AI applications can discover and access.
- How it is used
- Standardizes how services expose tools so agents can discover and invoke them without maintaining a custom wrapper for each service.
- How deep it goes
- Explains host, client, server, transport, discovery, and testing workflows; defers full protocol and security details to the specification.
- Examples
- Uses an AccuWeather server example and a workflow mixing a local calculation, an MCP-served query, and local formatting.
- What you can do after
- Apply it via FastMCP 2 or the TypeScript MCP SDK to expose tools, test with MCP Inspector, and connect agents.
Architecting an Apache Iceberg Lakehouse
The Model Context Protocol is a standardized protocol that lets large language models interact with external data, tools, and services through governed calls.
- How it is used
- Enables AI-native consumption of governed lakehouse data, tools, metrics, and workflows through standardized calls.
- How deep it goes
- Explains architecture, interactions, examples, and safeguards, but stops short of specifying server implementation details.
- Examples
- Uses dataset discovery, natural-language querying, governed metric retrieval, forecasting, and downstream workflow triggering.
- What you can do after
- Operate/configure AI agents that discover, query, and act on governed lakehouse capabilities through MCP servers.
Read how Architecting an Apache Iceberg Lakehouse explains MCP →
Building Reliable AI Systems
A protocol for connecting AI applications and agents to external data sources and tools.
- How it is used
- Connects AI agents to external APIs, databases, and services through standardized tool interfaces.
- How deep it goes
- Works through server and client roles, schemas, discovery, implementation, and error handling; leaves tool execution and security implementation to developers.
- Examples
- Uses a hiking-boot product search, an email-tool schema, and a locally running HTTP product-search tool.
- What you can do after
- Build it from scratch: implement Python tools with FastMCP and expose them for model discovery through an MCP server.
Financial AI in Practice
MCP is a protocol for standardizing how AI models or agents interact with external tools and data sources.
- How it is used
- Standardizes how financial AI agents access approved tools and shared data assets instead of maintaining separate function-calling integrations.
- How deep it goes
- Explains the integration problem and sketches tool discovery in conceptual pseudocode, but omits runnable client configuration and server details.
- Examples
- Contrasts manual stock-chart and community-search wrappers with a conceptual MCP client, then considers controlled access to a retention agent’s user-behavior mart.
- What you can do after
- Evaluate/understand it as one way to give agents controlled tool access, while recognizing that internal registries or service layers are alternatives.
Spring AI in Action
The Model Context Protocol is an open protocol for connecting AI applications with external tools and sources of contextual information.
- How it is used
- Connects generative AI applications with reusable collections of external tools and contextual resources.
- How deep it goes
- Defines servers, clients, tool invocation, and transports, while deferring Streamable HTTP support in Spring AI 1.0.0.
- Examples
- Uses integrations such as PostgreSQL, GitHub, Google Maps, and Claude Desktop configurations.
- What you can do after
- Apply MCP via Spring AI to reuse external tools in AI applications and extend LLM functionality.
Investing for Programmers
A model context protocol is a standardized interface that lets AI agents connect to external data sources and retrieve data in a consistent format.
- How it is used
- Connects AI agents to external data during response generation.
- How deep it goes
- Defines the interface, compares it with RAG and web search, and stops short of implementation details.
- Examples
- Uses an agent querying the latest stock share price while generating an answer.
- What you can do after
- Apply it via an MCP interface to let an LLM retrieve current external data during answer generation.
Introduction to Generative AI, Second Edition
The Model Context Protocol is an open standard for secure, two-way connections between external data sources, services, and AI-powered tools.
- How it is used
- Connects AI agents with external data and services through standardized server-client interactions.
- How deep it goes
- Explains servers, clients, data functions, prompt templates, and executable tools, but not implementation details.
- Examples
- Uses Splunk log queries, time-based log filtering, and automated log-irregularity analysis as concrete examples.
- What you can do after
- Apply it via MCP clients to request information and actions from services that provide MCP servers.
Read how Introduction to Generative AI, Second Edition explains MCP →
Authorization in Action
Model Context Protocol is a standard protocol for connecting AI applications to external tools and data sources.
- How it is used
- Connects AI agents with external tools and data through discoverable, structured tool invocations.
- How deep it goes
- Defines tool exposure, schemas, and server-side translation, but leaves authorization decisions to separate policy mechanisms.
- Examples
- Uses document search, calendar access, database queries, workflow actions, and an ACME document-search API.
- What you can do after
- Apply MCP to expose and invoke external capabilities without writing custom integration code for every system.
Building Distributed Applications that Work
Model Context Protocol is a protocol for connecting language-model applications with external tools, data sources, and capabilities through a standardized interface.
- How it is used
- Connects coding agents to current Aspire information while they work on an existing application.
- How deep it goes
- Defines the relationship between MCP and agentic tool calling, but does not specify MCP configuration or protocol details in the supplied excerpt.
- Examples
- Uses an Aspire setup and a refactoring scenario involving file inspection, code changes, and repeated test runs.
- What you can do after
- Evaluate/understand why a coding agent needs current external information and how tool results support iterative changes.
Read how Building Distributed Applications that Work explains MCP →
Building LLM Applications with DSPy
Model Context Protocol is a standard protocol for connecting language-model applications with external tools and data sources.
- How it is used
- Connects an agentic application to external services, including retrieval tools exposed by an MCP server.
- How deep it goes
- Defines host, client, server, and stdio/HTTP connections, stopping short of a broader protocol treatment.
- Examples
- Uses a host connected to three MCP servers and a server exposing WixQA retrieval methods.
- What you can do after
- Understand how a host uses clients to communicate with servers and execute tools.
Related concepts
- toolconfigures · A tool is an instrument or software capability that enables a person or system to perform a task, obtain information, or interact with resources.
- contractcontrasts with · An API contract is a specification of the operations, data, and behavioral expectations that govern interactions between software components.
- A2Acontrasts with · A2A is an interoperability protocol that enables software agents to discover, communicate, and coordinate with one another across systems.
- client-server architectureuses · A distributed computing pattern in which clients request services or resources from servers, which process the requests and provide results.
- N×M integration problemaddresses · The N×M integration problem is the growth in custom integrations that results when each of N applications must connect independently to each of M external systems or providers.
- schemauses · A data schema is a formal specification of the structure, fields, types, relationships, and constraints of data.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.