MCP Prompts
Prompts are message templates an MCP server offers for the user to pick, often surfaced as slash commands. The client lists them and fetches one, with arguments, when the user chooses it.
Client support
| Client | List and get a prompt | Tested |
|---|---|---|
| Claude Code v2.1.296 | Supported |
Required or optional?
Optional for servers: the spec only requires a server that offers this feature to declare the capability. Neither revision says whether a client must support it.
What the 2026-07-28 spec says
The Model Context Protocol (MCP) provides a standardized way for servers to expose prompt templates to clients. Prompts allow servers to provide structured messages and instructions for interacting with language models. [source]
This refers to who decides when the prompt is used, not who authors its content. Prompt content is defined by the server.
[source]
Servers that support prompts MUST declare the prompts capability in their DiscoverResult
[source]
Servers MAY also respond to prompts/get with an InputRequiredResult
[source]
What the 2025-11-25 spec says
The Model Context Protocol (MCP) provides a standardized way for servers to expose prompt templates to clients. Prompts allow servers to provide structured messages and instructions for interacting with language models. [source]
Prompts are designed to be user-controlled, meaning they are exposed from servers to clients with the intention of the user being able to explicitly select them for use.
[source]
Servers that support prompts MUST declare the prompts capability during initialization
[source]
Protocol methods
prompts/listprompts/getnotifications/prompts/list_changed