auth-test-mcp-remote
Remote MCP server for verifying API key authentication header propagation and debugging inbound HTTP headers.
README
auth-test-mcp-remote
Remote MCP server for verifying API key authentication header propagation.
Deployable to Render.com (free tier).
Purpose
Part of the custom API key header epic test suite.
Validates that WXO forwards the correct header name and value when a connection
with api_key_auth (and a custom label) is linked to a remote MCP toolkit.
Tools
| Tool | Auth required | What it returns |
|---|---|---|
ping |
No | Server name + API_KEY_HEADER_NAME currently configured |
get_api_key_variables |
Yes | Env vars containing API, KEY, or TOKEN |
echo_headers |
Yes | All inbound HTTP request headers — primary verification tool |
echo_query_and_cookies |
Yes | Inbound query params and cookies (for cookie/query location tests) |
Environment variables
| Variable | Default | Description |
|---|---|---|
API_KEY_HEADER_NAME |
x-api-key |
Header name the server expects in inbound requests |
EXPECTED_API_KEY |
(empty) | If set, validates the incoming key strictly — leave blank to accept any key |
HOST |
0.0.0.0 |
Bind address |
PORT |
8000 |
Bind port (Render injects 10000 automatically) |
Endpoints
| Path | Description |
|---|---|
/sse |
SSE transport (for WXO connection) |
/messages/ |
MCP post-handshake POST endpoint |
/health |
Health check (no auth) |
Deploying to Render.com (free tier)
Step 1 — Put the server in its own GitHub repository
Render deploys from a Git repo. The easiest approach is to push just this folder as a standalone repository.
cd agentic_data/test_case_files/z_api_key_custom_header/mcp_servers/auth_test_mcp_remote
git init
git add .
git commit -m "feat: auth-test-mcp-remote initial commit"
# Create a new repo on github.com first, then:
git remote add origin https://github.com/YOUR_USERNAME/auth-test-mcp-remote.git
git branch -M main
git push -u origin main
Step 2 — Create a Web Service on Render
- Go to dashboard.render.com → New + → Web Service
- Connect your GitHub account and select the
auth-test-mcp-remoterepository - Render will auto-detect
render.yamland pre-fill:- Environment: Python
- Build command:
pip install -e . - Start command:
auth-test-mcp-remote - Port: 10000
- Set the
EXPECTED_API_KEYsecret in the Environment tab (the value WXO will send in the header) - Click Create Web Service
Render assigns a public HTTPS URL like:
https://remote-mcp-server-sk0z.onrender.com
Step 3 — Verify the deployment
# Health check
curl https://remote-mcp-server-sk0z.onrender.com/health
# Expected:
# {"status":"healthy","service":"auth-test-mcp-remote","version":"0.1.0",
# "implementation":"FastMCP","api_key_header":"x-api-key"}
Free tier note: Render spins down free services after 15 min of inactivity. The first request after sleep takes ~30–60 s. WXO's 30 s import timeout may trip during a cold start — hit the health endpoint once to warm it up before running
orchestrate toolkits add.
Importing into WXO (after deployment)
Option A — orchestrate toolkits add (one-liner)
# Create and configure the connection first
orchestrate connections add -a auth_test_remote
for env in draft live; do
orchestrate connections configure -a auth_test_remote --env $env --type team --kind key_value
orchestrate connections set-credentials -a auth_test_remote --env $env \
-e "EXPECTED_API_KEY=<your-secret-key>"
done
# Import the toolkit
orchestrate toolkits add \
--kind mcp \
--name auth_test_mcp_remote \
--description "Remote MCP server for API key authentication testing" \
--url "https://remote-mcp-server-sk0z.onrender.com/sse" \
--transport sse \
--tools "*" \
--app-id auth_test_remote
Option B — Import from YAML file
Create auth_test_mcp_remote.yaml:
spec_version: v1
kind: mcp
name: auth_test_mcp_remote
transport: sse
server_url: https://remote-mcp-server-sk0z.onrender.com/sse
tools:
- "*"
connections:
- auth_test_remote
Then run:
orchestrate toolkits import -f auth_test_mcp_remote.yaml -a auth_test_remote
Running locally
pip install -e .
# Default header (x-api-key), no strict validation
auth-test-mcp-remote
# Custom header + strict validation
API_KEY_HEADER_NAME=x-custom-auth EXPECTED_API_KEY=mysecret auth-test-mcp-remote
# Or with Docker
docker-compose up
After deploying — update mcp_config_example.json
{
"mcpServers": {
"auth-test-remote": {
"url": "https://remote-mcp-server-sk0z.onrender.com/sse",
"headers": {
"x-api-key": "<your-secret-key>"
}
}
}
}
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。