Horizon MCP Auth Proxy

Horizon MCP Auth Proxy

A generic single-upstream MCP authentication proxy that wraps remote Streamable HTTP MCP servers behind Horizon OAuth, enabling clients to connect securely without carrying upstream API keys. It adds a fixed authentication header to upstream requests and supports bearer tokens, API keys, and custom headers.

Category
Visit Server

README

Horizon MCP Auth Proxy

这是一个可部署到 Prefect Horizon 的通用单上游 MCP 认证代理。它将原本需要 固定 HTTP 请求头的远程 Streamable HTTP MCP,包装为由 Horizon OAuth 保护的 MCP 地址。

MCP 客户端 -- Horizon OAuth --> 本代理 -- 固定认证请求头 --> 远程 MCP

上游 MCP 不需要在本地或本仓库中运行。本仓库中的 main.py:mcp 本身就是 Horizon 所要求的 FastMCP 服务实例。

环境变量

代理只需要三个环境变量:

变量 说明 示例
MCP_UPSTREAM_URL 完整的远程 Streamable HTTP MCP 地址 https://example.com/mcp
MCP_AUTH_HEADER_NAME 上游认证请求头名称 Authorization
MCP_AUTH_HEADER_VALUE 完整请求头值 Bearer your-token

MCP_AUTH_HEADER_VALUE 应在 Horizon 中作为 Secret 保存,不要提交到 GitHub。

Bearer Token 示例

MCP_UPSTREAM_URL=https://example.com/mcp
MCP_AUTH_HEADER_NAME=Authorization
MCP_AUTH_HEADER_VALUE=Bearer your-upstream-token

Google Stitch 示例

MCP_UPSTREAM_URL=https://stitch.googleapis.com/mcp
MCP_AUTH_HEADER_NAME=X-Goog-Api-Key
MCP_AUTH_HEADER_VALUE=your-google-api-key

Horizon 部署

  1. 将本仓库推送到 GitHub,并在 Horizon Deploy 中选择它。
  2. Entrypoint 填写 main.py:mcp
  3. 开启 Horizon Authentication。Horizon 将负责客户端侧 OAuth。
  4. 可以先不填写上游环境变量并完成第一次部署。此时服务是一个合法但没有工具的 Horizon MCP Auth Proxy (Unconfigured)
  5. 在部署设置中添加上述三个环境变量。
  6. 重新部署或重启服务。环境变量在进程启动时读取,不会热更新。
  7. 在 Horizon Inspector 中连接服务并确认上游工具已经出现。

部署后的客户端只连接:

https://<server-name>.fastmcp.app/mcp

客户端不需要、也不应该携带上游 API Key。代码会禁用下游请求头转发,因此 Horizon OAuth 的 Authorization Bearer Token 不会发送到上游。

如果只配置了部分环境变量,服务会启动失败并明确列出缺失的变量。这可以避免 代理在错误凭据状态下运行。要恢复“待配置”状态,需要同时删除三个变量并重新部署。

支持范围

  • 一个远程 Streamable HTTP MCP 上游
  • 一个固定认证请求头
  • Authorization: Bearer ...、API Key 和其他自定义认证头
  • Horizon 管理的下游 OAuth

当前不支持 SSE-only、stdio、多上游、上游 OAuth 登录流程或配置热更新。

本地运行

安装依赖:

python -m pip install -r requirements.txt

未配置状态可以直接启动:

fastmcp run main.py:mcp --transport http --port 8000

完整配置可以复制 .env.example.env,填写真实值后启动。FastMCP 默认加载 项目目录中的 .env,但真实 .env 已被 .gitignore 排除。

测试

python -m pip install -r requirements-dev.txt
ruff check main.py tests
mypy --strict main.py
python -m pytest -q
python -m compileall main.py tests
fastmcp inspect main.py:mcp

fastmcp inspect 在未配置状态下应成功,并显示一个没有上游工具的有效 MCP 服务。

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured