Introduction

What is Layer

Layer gives you everything you need to make your product usable by AI agents—whether they’re running in Claude, Cursor, VS Code, or any other MCP client.

We help you create and maintain a Model Context Protocol (MCP) server for your product so you can expose context, capabilities, and guidance to agents with minimal overhead. This includes support for resources, prompts, and tools—the key components MCP-powered agents rely on to understand and use your product effectively.

---

What’s Included

Layer provides:

• An easy-to-use interface for creating and managing your MCP server
• Support for:
  • Resources – Documentation or other structured context
  • Prompts – Instructional content or workflows that shape how agents respond
  • Tools – Structured API endpoint calls agents can use to interact with your product
• Automatic updates to keep your server aligned with the latest client-supported features
• An automatically generated Demo MCP Server to help you instantly test your sever's capabilities
• A CLI for generating your own self-hosted and customizable MCP server

---

Why MCP (and Why Layer)

The Model Context Protocol (MCP) standardizes how AI agents consume external context, call tools, and make decisions. It’s powerful—but still evolving. Many client platforms only support a subset of the spec, and behavior can vary across tools.

Layer makes creating your server and keeping your MCP server up to date with the latest client expectations super easy. We handle compatibility and protocol changes so you don’t have to.

---

What You’ll Find in These Docs

• How to use Layer's Management Dashboard to create and manage:
  • Resources (reference content for agents)
  • Prompts (agent workflows and behavior shaping)
  • Tools (structured API endpoint calls usable by agents)
• How to test your server inside supported agent clients
• How to generate and publish your own self-hosted MCP server
• Best practices for making your product easy to use via AI agents