MCP HTTP Client Server
A powerful MCP server for making HTTP requests, GraphQL queries, and TCP/Telnet connections from AI assistants.
README
<p align="center"> <img src="mcp-http.png" alt="MCP HTTP Client" title="MCP-HTTP-CLIENT" align="right" width="64"/> </p>
MCP HTTP Client Server
A powerful Model Context Protocol (MCP) server for making HTTP requests, GraphQL queries, and TCP/Telnet connections from AI assistants. Inspired by Postman but designed for AI-native workflows with enhanced response formatting, intelligent caching, and multiple content type support.
Why Use This?
- Seamlessly integrates with AI assistants like Claude
- 13.7x faster with intelligent LRU caching
- Auto-formatted JSON, status emojis, performance metrics
- JSON, form-data, URL-encoded - all supported
- very request shows response time
- atural language commands, no complex setup
Installation
Choose your preferred installation method:
Option 1: npm (Recommended for easy setup)
npm install -g @mcp/http-client
Option 2: Homebrew (macOS/Linux)
brew tap ferPrieto/mcp-http-client
brew install mcp-http-client
Option 3: GitHub Packages (for Kotlin/JVM projects)
Add the repository and dependency to your build.gradle.kts:
repositories {
maven {
url = uri("https://maven.pkg.github.com/ferPrieto/MCP-Http-Client")
credentials {
username = project.findProperty("gpr.user") as String? ?: System.getenv("GITHUB_ACTOR")
password = project.findProperty("gpr.key") as String? ?: System.getenv("GITHUB_TOKEN")
}
}
}
dependencies {
implementation("ferprieto.mcp:httpclient:1.0.0")
}
Note: GitHub Packages requires authentication. Generate a Personal Access Token with read:packages scope.
Option 4: Build from Source
git clone https://github.com/ferPrieto/MCP-Http-Client.git
cd MCP-Http-Client
./gradlew clean build
This generates build/libs/mcp-http-client-all.jar.
Configuration
Add to your MCP client configuration file:
- Cursor:
~/.cursor/mcp.json - Claude Desktop:
~/Library/Application Support/Claude/claude_desktop_config.json
If installed via npm:
{
"mcpServers": {
"http-client": {
"command": "npx",
"args": ["@mcp/http-client"]
}
}
}
If installed via Homebrew:
{
"mcpServers": {
"http-client": {
"command": "mcp-http-client"
}
}
}
If using JAR directly:
{
"mcpServers": {
"http-client": {
"command": "java",
"args": ["-jar", "/path/to/mcp-http-client-all.jar"]
}
}
}
Comparison with Postman
| Feature | MCP HTTP Client | Postman |
|---|---|---|
| Interface | 🤖 Natural Language | 🖱️ GUI |
| Setup Time | ⚡ 1 minute | ⏱️ 5+ minutes |
| HTTP Methods | ✅ All | ✅ All |
| GraphQL | ✅ Native | ✅ Yes |
| TCP/Telnet | ✅ Yes | ❌ No |
| Content Types | ✅ JSON, Form, URL-encoded | ✅ Many |
| Response Formatting | ✅ Auto pretty-print | ✅ Yes |
| Performance Cache | ✅ 13.7x faster | ❌ No |
| Response Time | ✅ Auto-tracked | ✅ Yes |
| Cost | 🆓 Free & Open Source | 💰 Free/Paid |
| AI Integration | ✅ Native | ❌ Manual |
Advanced Features
Content-Type Auto-Detection
The server automatically:
- Detects JSON responses and pretty-prints them
- Sets appropriate
Content-Typeheaders based onbodyType - Handles form-data with proper multipart boundaries
- URL-encodes form parameters automatically
Performance Metrics
Every request automatically tracks:
- Total request duration
- Response size
- Status codes
- Timing information
Roadmap
Future features planned for upcoming releases:
- Collections: Save and organize API requests (like Postman collections)
- Environments: Manage variables across different environments (Dev, Staging, Production)
- Authentication Helpers: Built-in support for Basic Auth, Bearer Token, API Key, OAuth2
- Request Chaining: Use response values in subsequent requests with variable substitution
- Postman Import: Import existing Postman collections for easy migration
License
MIT License
Made with ❤️ for the AI-native development workflow. Simpler than Postman, faster than manual curl commands, perfect for AI assistants!
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。