Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
作者提供的说明(英文原文)
Agentic AI quality engineering: run, debug and triage tests on real browsers and devices
这三项永远不会合并成一个分数:意见、测量和人气是不同的东西。
我们还没有为它写通俗的介绍。下面是我们从注册中心读取到的事实。
选择你使用的 AI 应用,然后复制设置。
你需要
你用的是哪个应用?
在上方选择你的应用。我们会记住它,下次按钮会显示“安装到 …”。
适用于网页版、桌面版和手机版 Claude:
https://mcp.lambdatest.com/mcp在终端中运行这条命令
claude mcp add --transport http lambdatest https://mcp.lambdatest.com/mcp粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"lambdatest": {
"url": "https://mcp.lambdatest.com/mcp"
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
粘贴到项目中的 .vscode/mcp.json
{
"servers": {
"lambdatest": {
"type": "http",
"url": "https://mcp.lambdatest.com/mcp"
}
}
}浏览器会先询问是否打开应用,然后应用会显示设置并请你确认。密钥的值会留空,由你自己填写。
ChatGPT 只能连接互联网上的服务器:
https://mcp.lambdatest.com/mcp菜单名称可能因套餐和版本不同而略有差异。
你的选择通过 Cookie 保存在这台设备上,无需账户。
这些设置是根据作者在注册中心的条目生成的。如果 README 的写法不同,请以 README 为准。
我们还没有读取这个服务器的工具列表;我们会轮流检查各个端点。
我们检查的是任何人都能看到的信息;我们没有阅读代码。不适用的检查项不计入总数。 · 检查于 2026年11月7日
注册名称已通过 GitHub 账户 LambdaTest 验证,该账户拥有此仓库
我们还没有从 GitHub 读取这个仓库
我们还没有从 GitHub 读取这个仓库
尚未检查;我们会轮流检查各个端点
尚未检查;我们会轮流检查各个端点
尚未检查;我们会在检查端点时读取工具列表
与注册中心里这个分类中所有活跃的服务器比较。只有当发布者拥有该仓库时,星标数才计入。
它在 1 项适用的基础检查中通过了 1 项。我们没有阅读代码。在添加密钥或令牌之前,请先看看这个服务器要求哪些权限,并使用只具备所需权限的密钥。
注册中心没有列出必需的密钥,不过有些服务器会在首次使用时要求你登录。
它在作者的云端运行。通过互联网连接,无需安装,但你发送的内容会经过作者的服务器。
GitHub 仓库最后更新于–,最新版本(1.0.7)于17天前进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,所以接受远程 MCP 服务器的应用也可以使用,例如开启开发者模式的 ChatGPT。
同一分类中有:Spec Driven Development, Remote Desktop Commander, SuperCollider
同一分类中其他作者的服务器,按星标数排序。
如何看卡片:★ GitHub 星标i三大支柱:评价 · 检查 · 使用情况i更新 = 最新版本发布到注册中心的时间i
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
Make sound with SuperCollider: synths, drums, chords, melodies and soundscapes over OSC.
按原文从作者的仓库复制。PopNx MCP 没有审查或背书其内容。HTML、图片和脚本已被移除。链接会离开本站。
Run, debug, and triage tests using natural language, directly from your IDE.
The TestMu AI MCP Server connects any MCP-compatible client to the TestMu AI (Formerly LambdaTest) platform — so you can orchestrate test runs, debug failures, analyze visual regressions, and run accessibility audits without leaving your editor or switching between dashboards.
This repository is the setup guide and reference for the hosted MCP server. The server itself runs as a managed service at
https://mcp.lambdatest.com/mcp— there is nothing to install or self-host.
| Tool | What it does |
|---|---|
| HyperExecute | AI-native test orchestration. Generates HyperExecute YAML and test runner commands, triggers jobs, and reports status — no manual YAML authoring. |
| Automation | Triages test failures by pulling execution details, command logs, network logs, and console errors into one place for root-cause analysis. |
| SmartUI | Explains visual regressions in plain English — pixel, layout, DOM, and perceptual differences, summarized rather than eyeballed. |
| Accessibility | Runs WCAG and a11y audits against hosted URLs or local React apps, returning violations with remediation guidance. |
| Test Manager | Covers the test management lifecycle — AI test case generation through to bulk result recording. |
Authentication is handled over OAuth. On first connection, your client opens a browser window to authorize against testmuai.com; no API keys need to be stored in config files.
The server endpoint is:
https://mcp.lambdatest.com/mcp
Add to ~/.cursor/mcp.json (global) or .cursor/mcp.json (per project):
{
"mcpServers": { "testmu-ai": { "url": "https://mcp.lambdatest.com/mcp" } } }
claude mcp add --transport http testmu-ai https://mcp.lambdatest.com/mcp
Or install from the TestMu AI plugin marketplace (Claude Code v2.1.292 or later):
claude plugin install kane-cli --marketplace LambdaTest/mcp-server-guide
claude plugin install rook --marketplace LambdaTest/mcp-server-guide claude plugin install testmu-ai --marketplace LambdaTest/mcp-server-guide
| Plugin | What you get |
|---|---|
kane-cli | Kane CLI skill for browser automation, mobile app testing and AI test authoring |
rook | Rook inside Claude Code for testing the AI agents you build |
testmu-ai | This MCP server plus the Kane CLI and Rook CLI skills |
On older versions, add the marketplace first with claude plugin marketplace add LambdaTest/mcp-server-guide, then run claude plugin install kane-cli@testmu-ai. Inside a Claude Code session, use /plugin marketplace add LambdaTest/mcp-server-guide and /plugin install kane-cli@testmu-ai.
Add to your MCP configuration:
{
"servers": { "testmu-ai": { "type": "http", "url": "https://mcp.lambdatest.com/mcp" } } }
Open the Cline panel, click the MCP Servers icon, then Configure > Remote Servers:
Or click Configure MCP Servers in Cline to open cline_mcp_settings.json and add:
{
"mcpServers": { "testmu-ai": { "type": "streamableHttp", "url": "https://mcp.lambdatest.com/mcp" } } }
There is nothing to clone or install. Click Authenticate when it appears to complete the OAuth sign-in.
Add to ~/.config/opencode/opencode.json (global) or opencode.json in your project root:
{
"$schema": "https://opencode.ai/config.json", "mcp": { "testmu-ai": { "type": "remote", "url": "https://mcp.lambdatest.com/mcp", "oauth": {} } } }
Then run opencode mcp auth testmu-ai to sign in through your browser. opencode mcp list should show testmu-ai connected.
Add to ~/.config/kilo/kilo.jsonc:
{
"$schema": "https://app.kilo.ai/config.json", "mcp": { "testmu-ai": { "type": "remote", "url": "https://mcp.lambdatest.com/mcp" } } }
Then open Kilo Settings > Agent Behaviour > MCP Servers and click Sign in next to testmu-ai.
Clients that don't yet support remote HTTP transport can connect through mcp-remote:
{
"mcpServers": { "testmu-ai": { "command": "npx", "args": ["-y", "mcp-remote@latest", "https://mcp.lambdatest.com/mcp"] } } }
Cursor · Claude Code · Claude Desktop · GitHub Copilot · OpenAI Codex · Windsurf · Cline · Continue · Zed AI · JetBrains AI Assistant · Antigravity · OpenCode · Kilo Code
Any client implementing the Model Context Protocol should work. Clients with native remote HTTP support connect directly; the rest use the mcp-remote fallback above.
Once connected, ask your assistant things like:
Generate a HyperExecute YAML for my Playwright TypeScript tests and run it.
My last automation build failed — pull the console and network logs and tell me why.
Summarise the SmartUI visual differences in my latest build.
Run an accessibility audit on https://staging.example.com and list WCAG violations by severity.
Server doesn't appear after editing config Fully quit and reopen the client. Reloading the window is not enough for most clients to re-read MCP configuration.
Config not being picked up Validate the JSON. A trailing comma or mismatched brace will cause the client to silently skip the entry.
`npx` not found (STDIO fallback) Confirm Node.js is on your PATH — which npx on macOS/Linux, where npx on Windows. Some clients don't inherit your shell environment, so an absolute path to npx may be needed.
Authentication loop or expired session Clear the stored authentication for this server in your client's MCP settings and reconnect to re-trigger the OAuth flow.
Built by TestMu AI (formerly LambdaTest)