Agent Readiness

Model Context Protocol (MCP)

An open JSON-RPC protocol that lets AI agents talk to tools, resources, and prompts through a uniform server surface. MCP is the most direct expression of "an API designed for an agent" — every API surface that ships an MCP server scores higher on agent readiness because agents can call it through one shared transport.

GitHub

Documentation

📖
Documentation
https://modelcontextprotocol.io/specification

SDKs

📦
GitHubRepository
https://github.com/modelcontextprotocol/specification

Other Resources

🔗
GitHubOrganization
https://github.com/modelcontextprotocol

API entry from apis.yml

apis.yml Raw ↑ 
aid: agent-readiness:model-context-protocol
name: Model Context Protocol (MCP)
tags:
- Agent Tooling
- Agent Readiness
- JSON-RPC
- Standards
humanURL: https://modelcontextprotocol.io
baseURL: https://modelcontextprotocol.io
properties:
- url: https://modelcontextprotocol.io/specification
  type: Documentation
- url: https://github.com/modelcontextprotocol
  type: GitHubOrganization
- url: https://github.com/modelcontextprotocol/specification
  type: GitHubRepository
description: An open JSON-RPC protocol that lets AI agents talk to tools, resources, and prompts through
  a uniform server surface. MCP is the most direct expression of "an API designed for an agent" — every
  API surface that ships an MCP server scores higher on agent readiness because agents can call it through
  one shared transport.