MCP Integration
The Model Context Protocol (MCP) Integration connects ReadyAPI to AI assistants - including Claude, GitHub Copilot, ChatGPT, and Cursor - through natural-language interactions. It installs a local MCP server inside ReadyAPI that exposes your projects, test suites, and APIs to any MCP-compatible client.
Once connected, you can create projects, import API specifications, generate and run tests, diagnose and heal failures, and detect the impact of API changes - all from your AI assistant, without switching to the ReadyAPI UI.
What you can do
Capability | How it works |
|---|---|
Project creation and spec import | Create a project and import an API definition conversationally. |
Test suite generation | Generate test cases for every endpoint directly from the spec. |
Test execution | Run suites and retrieve per-step results from your AI client. |
Test healing | Diagnose failing steps and apply assertion fixes automatically. |
API change-impact analysis | Diff two spec versions and surface affected tests. |
End-to-end automation | Chain all of the above using five composable slash commands. |
In this section
Getting Started - Prerequisites, installation, and authentication setup
Connect an MCP Client - Configuration for VS Code, Cursor, Claude Desktop, Claude Code, and others
Available Tools - Complete reference for all MCP tools exposed by the plugin
Canned Prompts (Staged Workflow) - Four-stage change-detection and impact analysis workflow
Scope and Limitations - What the integration supports and what it does not
Configuration Reference - Server properties and defaults
What's Coming - The planned roadmap
Release history
The MCP Plugin releases independently of ReadyAPI on an approximate two-week cadence. For ReadyAPI release history, see Version History.
Version | Release date | What changed |
|---|---|---|
1.0.0 | July 20, 2026 | Initial release - test execution, failure diagnosis and healing, CRUD tools, REST Request step creation, Environments and Auth Profiles read tools, assertion management |