api-testing-mcp
Enables HTTP API testing (GET, POST, PUT, DELETE) via the Model Context Protocol, powered by Playwright.
README
<p align="center"> <img src="./documentation/screen-shots/mcp-logo.png" alt="MCP LOGO" width=400 height=400> </p>
API Testing Using MCP Server (End User Guide)
A ready-to-use Model Context Protocol (MCP) server for HTTP API testing, powered by Playwright. This guide helps you install, configure, and run the MCP server locally as an npm package. <br> <br>
📋 Table of Contents
- Overview
- Prerequisites
- Setup
- Configurations
- Usage
- Example scenarios and prompts for MCP server usage
- Upcoming enhancements
- Troubleshooting
- Support <br> <br>
🌟 Overview
The Playwright API MCP Server enables API testing via the Model Context Protocol. It supports GET, POST, PUT, and DELETE requests.
<br> <br>
📋 Prerequisites
- Node.js: v18.0.0 or higher.
- npm: Comes with Node.js.
- VS Code: MCP-compatible client.
- Github Copilot: Make sure the GitHub Copilot extension is installed and enabled in VS Code.<br> Watch the video to understand how to enable Github Copilot: https://www.youtube.com/watch?v=etP5wq4Zl20
- How to use Github Copilot for free Watch the Video: https://www.youtube.com/watch?v=X_Aet9ndh_Y <br> <br>
🚀 Setup
-
Create a Node.js Project in VS Code:
- Create a new folder with a meaningful name (e.g.
my-project). - Open the folder in Visual Studio Code.
- Launch the integrated terminal (via
View > Terminal). - Choose your preferred shell (e.g.
Git Bash, Command Prompt, or PowerShell). - Run the following command to initialise the project:
npm init -y - This will generate a package.json file with default settings <br>
- Create a new folder with a meaningful name (e.g.
-
Install the MCP Server Package
- Now, run the below command to install the mcp server package.
npm i @automate-io/api-testing-mcp@latest
- Now, run the below command to install the mcp server package.
<br>
⚙️ Configurations
-
Open your project folder in VS Code
-
If the
mcp.jsonfile does not already exist, create it inside theroot/.vscode/directory. If the file is already present, you can skip this step -
Add the
api-testing-mcpconfiguration under theserverskey in themcp.jsonfile, as shown below<img src="./documentation/screen-shots/mcp-client-config-v2.png" alt="MCP Client Config">
{ "servers": { "api-testing-mcp": { "type": "stdio", "command": "node", "args": ["node_modules/@automate-io/api-testing-mcp/dist/server.js"], "cwd": "${workspaceFolder}" } }, "inputs": [] }
- ❗Important Note:If you are working in a
monorepo, ensure theargsfield in yourmcp.jsonfile is correctly configured. Refer to the Monorepo MCP Configuration guide for detailed instructions.<br> If your test code resides in astandalonerepository, no changes are needed, you are good to go. <br> <br>
📖 Getting started with MCP testing
- Start the MCP Server
-
Click the highlighted start button in the mcp.json file as shown below to start the MCP server.
<img src="./documentation/screen-shots/server-start-button-v2.png" alt="Server Start Button"> <br>
- Input the prompt
- Open GitHub Copilot chat window in VS Code with Agent mode selected.
- Enter a prompt that outlines the API you wish to test. <br> <br>
✅ Example scenarios and prompts for MCP server usage
- Scenario 1 : To test all endpoints in one go.<br>
- Scenario 2 : To test a single endpoint. <br> <br>
⏭️ Upcoming enhancements
- Implement schema validations for response payloads
- Add reusable resources to support API testing workflows
- Introduce predefined prompts to streamline common scenarios
- Provide a sample HTML test report template to guide LLm about the layout and formatting
- Refactor existing logic and implement robust error handling to improve reliability and maintainability <br> <br>
🔧 Troubleshooting
-
Server not running?
Double-check your MCP client configuration. -
Module not found?
Reinstall the package:npm i @automate-io/api-testing-mcp@latest
<br>
🆘 Support
- Review this guide and your MCP client documentation.
- For issues, open GitHub Issues here.
- Consult Model Context Protocol Documentation.
<br>
👍 Enjoy exploring and validating your APIs !
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。