MCP Simple Server
MCP 简易服务器 (HTTP 网桥)
这是一个简化的 MCP (Model Context Protocol) 服务器,通过 HTTP POST 接口暴露工具。它允许远程 LLM 和代理通过隧道或直接访问安全地调用本地或特定功能。
🚀 技术配置
端口:
3030端点:
/mcp方法:
POST认证:
Authorization: Bearer senha-facil-123协议: JSON-RPC 2.0 (适配 HTTP)
Related MCP server: local-env-mcp
🧪 使用 CURL 进行远程测试
您可以使用以下命令从任何具有网络访问权限的机器(或通过 Wireguard VPN)验证服务器的运行情况。
1. 列出可用工具
curl -X POST http://10.7.0.1:3030/mcp \
-H "Authorization: Bearer senha-facil-123" \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","method":"tools/list","id":1}'2. 调用 'hello' 工具
curl -X POST http://10.7.0.1:3030/mcp \
-H "Authorization: Bearer senha-facil-123" \
-H "Content-Type: application/json" \
-d '{"jsonrpc":"2.0","method":"tools/call","params":{"name":"hello","arguments":{"name":"Fabão"}},"id":2}'♊ 配置 Gemini-CLI (远程)
为了让您本地机器或其他 VPS 上的 gemini-cli 访问此服务器,您需要在 Gemini CLI 的配置文件(通常位于 ~/.gemini/config.yaml 或通过命令)中进行配置。
由于此服务器使用直接 HTTP(而非标准的 stdio),您可以使用“包装器”,或者如果服务器已扩展,则配置 SSE 访问,或者在可用时使用 fetch 插件。
配置示例(假设使用 stdio-to-http 代理):
# No gemini-cli
gemini mcp add remote-server --command "npx @modelcontextprotocol/inspector http://10.7.0.1:3030/mcp"
在 ~/.gemini/settings.json 文件中
"mcpServers": {
"local-mcp": {
"url": "http://192.168.1.100:3030/mcp",
"headers": {
"Authorization": "Bearer senha-facil-123"
}
}
}
注意:请将 IP 替换为正确的地址(例如 VPN 的 10.7.0.1)。
🤖 配置 Claude Code
Claude Code(以及 Claude Desktop)通常期望一个本地命令。要访问此远程服务器:
编辑
claude_desktop_config.json文件:macOS:
~/Library/Application Support/Claude/claude_desktop_config.jsonLinux:
~/.config/Claude/claude_desktop_config.json
添加配置:
{
"mcpServers": {
"mcp-simple-server": {
"command": "curl",
"args": [
"-s",
"-X", "POST",
"http://10.7.0.1:3030/mcp",
"-H", "Authorization: Bearer senha-facil-123",
"-H", "Content-Type: application/json",
"-d", "{\"jsonrpc\":\"2.0\",\"method\":\"tools/call\",\"params\":<params>}"
]
}
}
}提示:为了实现更稳健的集成,建议使用将 stdio 转换为服务器 HTTP API 的 mcp-proxy。
💬 配置 ChatGPT (GPT Codex / Actions)
要将此服务器用作 ChatGPT 中的 Action:
前往 Custom GPTs -> Create a GPT -> Configure -> Create new action。
在 Authentication 字段中,选择 API Key,选择 Bearer 并输入:
senha-facil-123。在 Schema 字段中,使用 OpenAPI 规范(下方为简化示例):
openapi: 3.0.0
info:
title: MCP Simple Server
version: 1.0.0
servers:
- url: http://195.35.42.89:3030 # IP Público ou DNS
paths:
/mcp:
post:
operationId: callMcpTool
summary: Chama uma ferramenta do MCP
requestBody:
content:
application/json:
schema:
type: object
properties:
method: {type: string, example: "tools/call"}
params: {type: object}
id: {type: integer, example: 1}
responses:
'200':
description: Sucesso🐳 Docker
本项目已准备好在容器中运行。
1. 构建镜像
在本地生成 Docker 镜像:
docker build -t mcp-simple-server .2. 启动容器 (Docker Compose)
推荐的运行方式是使用随附的 docker-compose.yml,它配置了网络和环境变量:
docker compose up -d容器将以 mcp-simple-server 名称创建,并可在 3030 端口访问。
3. 检查状态和日志
# Ver se o container está rodando
docker ps | grep mcp-simple-server
# Ver logs do servidor
docker compose logs -f🛠️ 开发
要在本地运行服务器:
cd ~/apps/mcp-simple-server
npm install
node server.js服务器将在 http://localhost:3030 监听。
This server cannot be deployed
Maintenance
Related MCP Connectors
Zero-setup MCP gateway securely connecting AI to your tools with authentication and workflows
MCP server exposing the Backtest360 engine API as tools for AI agents.
Host your MCP tool over streamable HTTP in one command.
Nifty's MCP server — exposes tasks, projects, messages, and files as tools for AI agents.
Related MCP Servers
- FlicenseNot gradedqualityDmaintenanceEnables remote access to the MemPalace MCP server via HTTP, supporting bearer token authentication and concurrent clients while exposing all mempalace tools.-
- FlicenseNot gradedqualityDmaintenanceExposes your local machine's filesystem, git, shell, network, databases, and system to any MCP-compatible LLM client over HTTP.6-
- AlicenseNot gradedqualityAmaintenanceLifts local stdio MCP servers into remote Streamable HTTP endpoints for cloud-hosted AI clients, with bearer-token auth and tool policy filtering.9 npmMIT
- FlicenseNot gradedqualityCmaintenanceConverts any MCP server into simple HTTP endpoints, enabling AI agents that only support HTTP (like ChatGPT Custom GPT or web_fetch) to use MCP tools.-