linktest-mcp-server
Automate Google Play closed testing with AI - skip the hassle of finding testers and waiting 14+ days by telling your AI to handle it.
README
linktest-mcp-server
Automate Google Play closed testing with AI. Skip the hassle of finding 12+ testers and waiting 14+ days — just tell your AI to handle it.
The Problem
To publish on Google Play, you need:
- 12+ testers opted in to your closed testing track
- 14+ days of continuous testing
Finding testers, getting them to opt in, and keeping them active for 2+ weeks is the biggest bottleneck for indie developers.
The Solution
LinkTest runs your app on 12+ real Android devices for 14+ days — automatically. Just tell your AI:
"Run closed testing for my app com.example.myapp"
That's it. $5 per order. No subscriptions. No tester recruitment.
Quick Start
1. Get your API token
Sign up at linktest-prod.web.app and copy your token from Settings.
2. Install
Works with any MCP-compatible AI client — Claude, Cursor, Windsurf, VS Code (Copilot), Cline, Continue, and more.
<details open> <summary><strong>Claude Desktop</strong></summary>
Add to your config file:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json - Windows:
%APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"linktest": {
"command": "npx",
"args": ["-y", "linktest-mcp-server@latest"],
"env": {
"LINKTEST_API_TOKEN": "your-token-here"
}
}
}
}
</details>
<details> <summary><strong>Claude Code</strong></summary>
claude mcp add linktest -s user -e LINKTEST_API_TOKEN=your-token-here -- npx -y linktest-mcp-server@latest
</details>
<details> <summary><strong>Cursor</strong></summary>
Add to .cursor/mcp.json in your project root:
{
"mcpServers": {
"linktest": {
"command": "npx",
"args": ["-y", "linktest-mcp-server@latest"],
"env": {
"LINKTEST_API_TOKEN": "your-token-here"
}
}
}
}
</details>
<details> <summary><strong>Windsurf</strong></summary>
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"linktest": {
"command": "npx",
"args": ["-y", "linktest-mcp-server@latest"],
"env": {
"LINKTEST_API_TOKEN": "your-token-here"
}
}
}
}
</details>
<details> <summary><strong>Other MCP Clients</strong></summary>
Any client that supports the Model Context Protocol can use this server. The config is the same:
- Command:
npx -y linktest-mcp-server@latest - Environment:
LINKTEST_API_TOKEN=your-token-here</details>
3. Start testing
Just tell your AI:
"I need to run Google Play closed testing for my app. The package name is com.example.myapp."
The AI will guide you through the entire process automatically.
Tools
| Tool | Description |
|---|---|
linktest_testers |
Get tester email list to register in Play Console (always start here) |
linktest_validate |
Validate your opt-in link format before ordering |
linktest_verify |
Verify real devices can access your app |
linktest_order |
Place a testing order ($5) and get payment link |
linktest_status |
Check order status, progress, and days remaining |
linktest_screenshots |
View daily screenshots from test devices |
linktest_retry |
Free retry if rejected by Google Play (1 per order) |
linktest_cancel |
Cancel or complete an order early |
linktest_billing |
View payment history and receipts |
Workflow
The AI handles this automatically, but here's what happens:
1. linktest_testers → Get emails, register in Play Console
2. linktest_validate → Check opt-in link
3. linktest_verify → Confirm device access (takes ~30s)
4. linktest_order → Create order → pay $5
5. linktest_status → Monitor progress
6. linktest_screenshots → Visual confirmation
Example Prompts
| What you say | What happens |
|---|---|
| "Start closed testing for my app" | Gets tester emails and guides you through setup |
| "Run closed testing for com.example.app" | Full workflow from validation to order |
| "How's my testing going?" | Shows status of all active orders |
| "Show me screenshots" | Displays latest device screenshots |
| "My app was rejected, can I retry?" | Uses free retry to extend testing 14 more days |
| "Cancel my order" | Cancels or completes the order |
Pricing
$5 per order — one-time payment, no subscription.
Each order includes:
- 12+ real Android devices
- 14+ days of daily app execution
- Daily screenshots for verification
- Free retry if rejected by Google Play (1 per order)
- Automatic management — no manual work needed
How It Works
- You register LinkTest tester emails in your Google Play Console
- You pay $5 per app
- LinkTest installs your app on 12+ real Android devices
- Devices run your app daily for 14+ days
- Screenshots are captured daily so you can verify
- Done — your closed testing requirement is met
Troubleshooting
MCP not loading?
If npx is installed at a non-standard path (e.g., via nvm or .local), change "command": "npx" to the full path:
which npx # Find your npx path, then use it in the config
Better AI guidance
Add this to your Claude system prompt or CLAUDE.md for step-by-step order guidance:
When using LinkTest tools, always read the resource "linktest://workflow-guide" first to follow the correct order.
Requirements
- A Google Play Developer account
- An app with a closed testing track set up
- Node.js 18+
Environment Variables
| Variable | Required | Description |
|---|---|---|
LINKTEST_API_TOKEN |
Yes | Your API token from Settings |
LINKTEST_API_URL |
No | Custom API URL (default: production) |
Links
License
MIT
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。