This is v1.x of the MCP TypeScript SDK, the maintenance line. It implements the MCP spec up to 2025-11-25. Support for the 2026-07-28 spec is not planned for v1.x: it is in v2, the current stable release, which
ships as separate packages, and . See the
and the .
Staying on v1.x? It continues to receive bug fixes and security updates for at least six months after the v2 release (2026-07-27). The v1 documentation is at https://ts.sdk.modelcontextprotocol.io/.
The Model Context Protocol allows applications to provide context for LLMs in a standardized way, separating the concerns of providing context from the actual LLM interaction. This TypeScript SDK implements
the 2025-11-25 MCP specification, making it easy to:
Create MCP servers that expose resources, prompts and tools
Build MCP clients that can connect to any MCP server
Use standard transports like stdio and Streamable HTTP
Installation
npm install @modelcontextprotocol/sdk zod
This SDK has a required peer dependency on zod for schema validation. The SDK internally imports from zod/v4, but maintains backwards compatibility with projects using Zod v3.25 or later. You can use either API in your code by importing from zod/v3 or zod/v4:
Quick Start
To see the SDK in action end-to-end, start from the runnable examples in src/examples:
This pair of examples demonstrates tools, resources, prompts, sampling, elicitation, tasks and logging. For a guided walkthrough and variations (stateless servers, JSON-only responses, SSE compatibility, OAuth, etc.), see docs/server.md and
docs/client.md.
Core Concepts
Servers and transports
An MCP server is typically created with McpServer and connected to a transport such as Streamable HTTP or stdio. The SDK supports:
Streamable HTTP for remote servers (recommended).
HTTP + SSE for backwards compatibility only.
stdio for local, process-spawned integrations.
Runnable server examples live under src/examples/server and are documented in docs/server.md.
Tools, resources, prompts
Tools let LLMs ask your server to take actions (computation, side effects, network calls).
Resources expose read-only data that clients can surface to users or models.
Prompts are reusable templates that help users talk to models in a consistent way.
The detailed APIs, including ResourceTemplate, completions, and display-name metadata, are covered in docs/server.md, with runnable implementations in simpleStreamableHttp.ts.
Capabilities: sampling, elicitation, and tasks
The SDK includes higher-level capabilities for richer workflows:
Sampling: server-side tools can ask connected clients to run LLM completions.
Form elicitation: tools can request non-sensitive input via structured forms.
URL elicitation: servers can ask users to complete secure flows in a browser (e.g., API key entry, payments, OAuth).
Tasks (experimental): long-running tool calls can be turned into tasks that you poll or resume later.
Conceptual overviews and links to runnable examples are in:
The high-level Client class connects to MCP servers over different transports and exposes helpers like listTools, callTool, listResources, readResource, listPrompts, and getPrompt.
Runnable clients live under src/examples/client and are described in docs/client.md, including:
OAuth-enabled clients and polling/parallel examples
Node.js Web Crypto (globalThis.crypto) compatibility
Some parts of the SDK (for example, JWT-based client authentication in auth-extensions.ts via jose) rely on the Web Crypto API exposed as globalThis.crypto.
See docs/faq.md for details on supported Node.js versions and how to polyfill globalThis.crypto when running on older Node.js runtimes.
Examples
The SDK ships runnable examples under src/examples. Use these tables to find the scenario you care about and jump straight to the corresponding code and docs.
Server examples
Scenario
Description
Example file(s)
Related docs
Streamable HTTP server (stateful)
Feature-rich server with tools, resources, prompts, logging, tasks, sampling, and optional OAuth.