使用 Playwright 和 Lighthouse 分析网站性能的 MCP 服务器。通过模型上下文协议 (MCP)，LLM 可以执行网站性能分析。
使用 Lighthouse 进行性能分析
屏幕截图
Related MCP server: Playwright MCP Server
Node.js 18 或更高版本
npm
# Clone the repository
git clone https://github.com/kbyk004/playwright-lighthouse-mcp.git
cd playwright-lighthouse-mcp
# Install dependencies
npm install
npx playwright install
# Build
npm run buildnpm run inspector此服务器旨在与支持模型上下文协议 (MCP) 的客户端配合使用。例如，它可以与 Claude for Desktop 集成。
将以下内容添加到 Claude for Desktop 配置文件（ ~/Library/Application Support/Claude/claude_desktop_config.json ）：
{
"mcpServers": {
"playwright-lighthouse": {
"command": "node",
"args": [
"/path-to/playwright-lighthouse-mcp/build/index.js"
]
}
}
}对当前打开的页面运行 Lighthouse 性能分析。
参数：
url ：要分析的网站的 URL
categories ：要分析的类别数组（默认值：[“performance”]）
可用类别：“性能”、“可访问性”、“最佳实践”、“SEO”、“PWA”
maxItems ：每个类别显示的最大改进项目数（默认值：3，最大值：5）
截取当前打开的页面的屏幕截图。
参数：
url ：要捕获的网站的 URL
fullPage ：如果为 true，则捕获整个页面的屏幕截图（默认值：false）
分析结果包括：
每个选定类别的总体得分，带有颜色指示器
按类别分组的关键改进领域
保存的报告文件的路径
MIT 许可证 - 详情请参阅许可证
Runs a Lighthouse performance analysis on the currently open page
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | URL of the website you want to analyze | |
| categories | No | Categories to analyze (performance, accessibility, best-practices, seo, pwa) | |
| maxItems | No | Maximum number of improvement items to display for each category |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries full burden for behavioral disclosure. While it states what the tool does, it lacks critical behavioral information: whether this requires network access, how long the analysis might take, what format the results come in, whether it modifies the page, or any error conditions. The description is insufficient for a tool that performs external analysis.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, efficient sentence that states the core functionality without unnecessary words. It's appropriately sized for the tool's purpose and gets straight to the point.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a performance analysis tool with no annotations and no output schema, the description is incomplete. It doesn't explain what kind of results to expect, how they're formatted, whether the analysis is synchronous or asynchronous, or any limitations. The agent would be operating with significant uncertainty about the tool's behavior and outputs.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
With 100% schema description coverage, the schema already documents all three parameters thoroughly. The description adds no additional parameter semantics beyond what's in the schema. The baseline score of 3 reflects adequate parameter documentation through the schema alone.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('Runs a Lighthouse performance analysis') and the target ('on the currently open page'), providing a specific verb+resource combination. However, it doesn't explicitly distinguish this tool from its sibling 'take-screenshot' beyond the different functionality implied by the names.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives, nor does it mention any prerequisites or context for usage. There's no comparison with the sibling 'take-screenshot' tool or indication of when performance analysis versus screenshot capture would be appropriate.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Takes a screenshot of the currently open page
| Name | Required | Description | Default |
|---|---|---|---|
| url | Yes | URL of the website you want to capture | |
| fullPage | No | If true, captures a screenshot of the entire page |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden. It states the action but lacks behavioral details: it doesn't mention output format (e.g., image type, storage location), error conditions (e.g., invalid URL, timeout), or side effects (e.g., whether it opens a browser). For a tool with no annotation coverage, this is a significant gap in transparency.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, efficient sentence that directly states the tool's purpose without unnecessary words. It is front-loaded and appropriately sized, earning its place with zero waste.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's complexity (involves browser interaction and image capture), lack of annotations, and no output schema, the description is incomplete. It doesn't cover behavioral aspects like output format, error handling, or dependencies, which are crucial for an AI agent to use it correctly. The description should do more to compensate for these gaps.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents both parameters ('url' and 'fullPage') with clear descriptions. The description adds no additional parameter semantics beyond what's in the schema, such as explaining 'currently open page' in relation to the 'url' parameter. Baseline 3 is appropriate when the schema does the heavy lifting.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('takes a screenshot') and target ('currently open page'), providing a specific verb+resource. However, it doesn't explicitly differentiate from the sibling tool 'run-lighthouse', which might also involve page analysis but serves a different purpose. The description is unambiguous about what the tool does.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides no guidance on when to use this tool versus alternatives like 'run-lighthouse'. It mentions 'currently open page' but doesn't specify prerequisites (e.g., whether a browser must be active) or exclusions. Usage context is implied but not explicit, leaving gaps for an AI agent to infer.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
run-lighthousetake-screenshotScored across 2 tools
The two tools have clearly distinct purposes: one runs a performance analysis (run-lighthouse) and the other captures a visual snapshot (take-screenshot). There is no overlap in functionality, making it easy for an agent to select the correct tool based on the task.
Both tools follow a consistent verb_noun pattern (run-lighthouse, take-screenshot), using hyphenated lowercase names. This predictable naming scheme enhances readability and usability for agents.
With only two tools, the server feels under-scoped for a Playwright-Lighthouse integration, which typically involves more operations like navigation, interaction, or detailed reporting. This limited set may hinder complex workflows that agents might expect from such a domain.
The tool surface is severely incomplete for browser automation and performance testing. Missing are essential operations like opening/closing pages, interacting with elements, running audits with specific configurations, or exporting results, which are core to the domain and could lead to agent failures in broader tasks.
SEO MCP server: crawl your site, find AI-visibility gaps, and ship the fix from your coding agent.
Analyze web performance and get optimization insights from GTmetrix, directly in your AI workflow.
Enable language models to perform advanced AI-powered web scraping with enterprise-grade reliabili…
Free technical-SEO audit MCP: crawl a site, run checks, return an LLM-ready shareable report.
