NestJS FAQs

Question 1

Can I access and execute tools or prompts via HTTP?

Accepted Answer

Absolutely. The module provides dedicated HTTP endpoints to list all registered tools and prompts, execute tools with a JSON payload, and obtain generated prompt messages for LLMs, enhancing accessibility and usability.

Question 2

How do I register tools and prompts using this module?

Accepted Answer

Tools and prompts are registered declaratively using `@McpTool()` and `@McpPrompt()` decorators, respectively. The module automatically detects and exposes all registered providers within your NestJS application.

Question 3

Does NestJS MCP integrate with OpenAI Function Calls?

Accepted Answer

Yes, NestJS MCP offers seamless integration with OpenAI Function Calls. It automatically structures your defined tools to be compatible, making it easy to build AI agents that interact with your application's functionalities.

Question 4

What is the NestJS MCP Module?

Accepted Answer

The NestJS MCP (Model Context Protocol) module empowers your NestJS applications to define, expose, and manage 'tools' (functions) and 'prompts' for Large Language Models (LLMs) or HTTP, complete with automatic detection and validation.

Question 5

How does NestJS MCP handle input validation for tools?

Accepted Answer

NestJS MCP includes built-in input data validation for tools. You define an `inputSchema` for each tool, and the module automatically validates incoming HTTP requests or LLM tool calls against this schema before execution, ensuring data integrity.

NestJS

NestJS

主要功能

使用案例

主要功能

使用案例