原型 · 预览版:收录官方注册中心里的全部服务器,每 30 分钟同步一次。使用 Google 登录即可撰写评价。原型 · 每 30 分钟更新
PopNx MCP为使用 AI 的人准备的 MCP 服务器 · 用大白话讲清楚 登录

PageLens AI

✓ 品牌同名 GitHub · github.com/pagelens-ai

作者:PageLens-AI版本 1.0.0i★ 1i在云端运行官方网站 ↗
注册名称io.github.PageLens-AI/pagelensai-mcp-server

安装方法

能做什么

AI website audit: security, SEO, performance, UX and accessibility checks with actionable fixes.

作者提供的原文

  • 连接方式 在云端运行
  • 账号或密钥 连接时登录该服务(OAuth)
  • 分类 网络安全

为什么排在这里

列表按使用分(满分 100)排序,由以下部分组成

热度4/80GitHub 星标或每周下载量:较大的计 60,较小的计 20
新鲜度3/1230 天内有发布或代码更新得满分
仍在响应8/8上次检查时云端端点有响应
使用分15/100

质量: 基础检查通过 5/5 项(仅展示,不用于排序) · 上升势头: 暂无数据 (只用于“上升中”标签)

安装

选择你使用的 AI 应用,然后复制设置。

用一句话让 AI 帮你安装

请帮我安装 PageLens AI。按照 https://mcp.popnx.com/i/pagelens-ai-pagelensai-mcp-server.md 操作,先把命令给我看,等我同意后再运行。

粘贴到能在你电脑上工作的 AI(如 Claude Code、Cursor)。它会读取我们的步骤,先给你看命令,等你同意。无需密钥,无需注册。 · 查看 AI 会读取的全文

通过 5/5 项检查 · 你的提问会经过开发者的服务器(www.pagelensai.com) · ✓ 命令已检查 2026年11月13日

第一步:你用哪个 AI?

选择你用的应用,步骤会显示在下方。

可连接 Claude · 约 2 分钟

不需要密钥,不用安装程序 · 在 claude.ai 或桌面应用设置一次,即可在聊天中使用 · 免费账户可添加 1 个 · Team/Enterprise 方案由所有者(Owner)添加

  1. 打开 Claude 的 Connectors 页面

    CustomizeConnectors

    CustomizeConnectors1Add2https://…3Add4
  2. 点击“Add” → “Custom” → “Web”,填写这 2 项

    把这里的内容复制到 Claude 中同名的输入框,然后点击 Continue

    Name · 名称PageLens AI
    Remote MCP server URL · 地址https://www.pagelensai.com/api/mcp

    在 Claude 检测到的登录设置页点击 Continue · Authentication 按作者说明选择(Sign in now、Sign in when needed 或 No sign in),然后点击 Add

  3. 在聊天中开启

    +ConnectorsPageLens AI

  4. 试一试

    用 PageLens AI 告诉我你能帮我做什么,然后做一件给我看

    ✓ 如果 Claude 说正在使用 PageLens AI,或回答中出现 PageLens AI 的工具名称,就表示连接成功

菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同

能否连接 ChatGPT 取决于方案

完整 MCP:网页版 ChatGPT Business、Enterprise 或 Edu · Pro:以读取/获取权限连接,修改数据的工具可能用不了 · Plus:官方文档尚未确认,如果在插件页看到添加自定义 MCP 的选项可以试试 · 仅限网页

  1. 在网页版 ChatGPT 打开添加 MCP 的页面

    Plugins+Add custom MCP server

    ChatGPTPlugins1+2https://…3Create4
  2. 填写名称和地址

    输入名称,在 Connection 下粘贴地址,勾选风险提示,然后点击 Create as a plugin

    Name · 名称PageLens AI
    Connection · 地址https://www.pagelensai.com/api/mcp
  3. 在聊天中选择 PageLens AI

    开启新聊天,输入前先从插件列表中选择 PageLens AI

  4. 试一试

    用 PageLens AI 告诉我你能帮我做什么,然后做一件给我看

    ✓ 如果 ChatGPT 说正在使用 PageLens AI,或回答中出现 PageLens AI 的工具名称,就表示连接成功

菜单名称已于 2026年11月11日 对照 OpenAI 帮助页面核实 · 你的界面可能因版本略有不同

可连接 Claude Code · 约 2 分钟

不需要密钥,不用安装程序

  1. 打开终端

    Mac:按 Command+空格,输入 Terminal · Windows:打开 PowerShell

  2. 粘贴这条命令并按 Enter

    claude mcp add --transport http --scope user pagelensai https://www.pagelensai.com/api/mcp
  3. 在 Claude Code 中输入 /mcp

    应能看到 PageLens AI 在列表中并显示已连接

  4. 试一试

    用 PageLens AI 告诉我你能帮我做什么,然后做一件给我看

    ✓ 如果 Claude Code 说正在使用 PageLens AI,或回答中出现 PageLens AI 的工具名称,就表示连接成功

菜单名称已于 2026年11月11日 对照 Anthropic 帮助页面核实 · 你的界面可能因版本略有不同

可连接 Cursor · 约 2 分钟

不需要密钥,不用安装程序

  1. 点击这个按钮打开 Cursor

    在 Cursor 中打开

    Cursor 会请你确认安装,点击 Install

  2. 在 Cursor 中以 Agent 模式打开 AI 聊天

  3. 试一试

    用 PageLens AI 告诉我你能帮我做什么,然后做一件给我看

    ✓ 如果 Cursor 说正在使用 PageLens AI,或回答中出现 PageLens AI 的工具名称,就表示连接成功

菜单名称已于 2026年11月11日 对照 Cursor 帮助页面核实 · 你的界面可能因版本略有不同

可连接 VS Code · 约 2 分钟

不需要密钥,不用安装程序

  1. 点击这个按钮打开 VS Code

    在 VS Code 中打开

    VS Code 会请你确认安装,点击 Install

  2. 在 VS Code 中以 Agent 模式打开 AI 聊天

  3. 试一试

    用 PageLens AI 告诉我你能帮我做什么,然后做一件给我看

    ✓ 如果 VS Code 说正在使用 PageLens AI,或回答中出现 PageLens AI 的工具名称,就表示连接成功

菜单名称已于 2026年11月11日 对照 VS Code 帮助页面核实 · 你的界面可能因版本略有不同

其他支持 MCP 的应用(如 Windsurf、Zed、Gemini CLI)添加方式各不相同,请在应用列表中查看

查看所有支持 MCP 的应用 →

通过网页连接的应用使用的地址

地址https://www.pagelensai.com/api/mcp

JSON 设置在下方“开发者专用”中

卡住了?问问正在用 PageLens AI 的人💬 讨论 →
开发者专用:终端命令和 JSON 设置

通过网页连接

Claude Code

claude mcp add --transport http --scope user pagelensai https://www.pagelensai.com/api/mcp

Cursor (~/.cursor/mcp.json)

{
  "mcpServers": {
    "pagelensai": {
      "url": "https://www.pagelensai.com/api/mcp"
    }
  }
}

VS Code (mcp.json)

{
  "servers": {
    "pagelensai": {
      "type": "http",
      "url": "https://www.pagelensai.com/api/mcp"
    }
  }
}

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。

一览

每周下载量i
不在 npm/PyPI 上
连接方式
在云端运行
传输方式
streamable-http
最新版本i
1.0.0
发布到注册中心i
2026年4月23日
代码最后更新i
2026年6月15日
最近检查
2026年11月16日
作者
PageLens-AI
注册状态
活跃

哪些应用可以使用它

依据这个服务器的连接方式和各应用自己的文档,检查于 2026年11月10日。

通过互联网连接,适用于

ClaudeChatGPTGooseLM StudioMsty StudioCherry StudioCursorVS Code + GitHub CopilotVisual StudioWindsurf (Devin Desktop)ZedKiroClineContinueJetBrains AI AssistantClaude CodeCodex CLIGemini CLIOpenCodenodecodeWarpLibreChatOpen WebUIRaycast

全部应用及其优缺点 →

它能做什么

服务器要求先登录,才会列出它的工具。

基础检查i

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20日

  • ✓ 通过

    作者与仓库一致

    注册名称已通过 GitHub 账户 PageLens-AI 验证,该账户拥有此仓库

  • ✓ 通过

    仍在维护

    代码最后更新于5个月前

  • ✓ 通过

    有许可证

    使用 MIT

  • ✓ 通过

    端点有响应

    在 406 毫秒内响应(HTTP 401)

  • ✓ 通过

    使用 OAuth 登录

    声明支持 OAuth 登录,你的 AI 应用可以带你去登录

  • … 尚未检查

    工具描述中没有隐藏指令

    尚未检查;我们会在检查端点时读取工具列表

我们检查什么 → · JSON

作者:申请重新检查或报告信息错误
给作者:在 README 中加徽章

如果你是作者,可把这段代码贴到 README。徽章会链接回本页。

[![Featured on PopNx MCP](https://mcp.popnx.com/badge/pagelens-ai-pagelensai-mcp-server.svg)](https://mcp.popnx.com/en/servers/pagelens-ai-pagelensai-mcp-server)

讨论

提问、分享技巧或报告无法使用。按“有帮助”数量排序。

还没有帖子,欢迎第一个提问或分享。

此处内容由用户撰写,不代表 PopNx 团队。

登录后发帖

用户评价

还没有评价

登录后写评价

与“网络安全”中的其他服务器比较i

  • 星标数多于此分类中其他 1120 个服务器的 74%
  • 发布到注册中心的时间比此分类中其他 1120 个服务器的 11% 更新
  • 基础检查结果等于或优于此分类中其他 1120 个服务器的 99%

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。

常见问题

PageLens AI 安全吗?

它在 5 项适用的基础检查中通过了 5 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。

它需要哪些密钥或代码?

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。

它在我的电脑上运行还是在云端?

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。

它还在维护吗?

GitHub 仓库最后更新于5个月前,最新版本(1.0.0)于7个月前进入注册中心。

哪些应用可以使用它?

支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。它通过 URL 连接,因此也能用于接受远程 MCP 的应用,例如网页版 ChatGPT(Business 及以上为完整 MCP;Pro 为只读)。

有什么可以替代它?

同一分类中有:GitHits, Dependency Management MCP Server, Snyk MCP, TokenSafe, Draugr

功能相近的其他选择

同一分类、其他作者、命令已通过检查。官方的排在前面,再按星标排序。

如何看卡片:↓ 每周下载量i★ 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

GitHits

Search public open-source code, documentation, metadata, vulnerabilities, changelogs, and examples.

作者:githits-com正在增长↓ 3807/周★ 138昨天更新

电脑或云端✓ 服务方 · githits.com

Snyk MCP

Easily find and fix security issues in your applications leveraging Snyk platform capabilities.

作者:snyk广泛使用↓ 76.2万/周★ 617个月前更新

在你的电脑上运行✓ 服务方 · snyk.io

TokenSafe

Your AI agent is smart enough to code all day. Local firewall & zero-loss runtime.

作者:tokensafe-dev广泛使用↓ 6785/周★ 47今天更新

电脑或云端✓ 服务方 · tokensafe.dev

Draugr

Security scanning for AI agents: SAST, SCA, secrets, IaC, DAST, ranked by real risk.

作者:draugr-dev起步阶段★ 303天前更新

在你的电脑上运行✓ 服务方 · draugr.dev
与相似项目对比
本项
PageLens AI
GitHitsDependency Management MCP ServerSnyk MCPTokenSafeDraugr
运行位置在云端运行,无需安装云端或安装在电脑上都可以在云端运行,无需安装需要安装在电脑上云端或安装在电脑上都可以需要安装在电脑上
需要密钥否否否否否否
可在 Claude 中按 + 添加否否否否否否
基础检查5/55/53/43/35/53/3
许可证MITApache-2.0–Apache-2.0MITApache-2.0
价格登记中未列出费用登记中未列出费用登记中未列出费用登记中未列出费用登记中未列出费用登记中未列出费用

作者的 README

打开 README

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。

PageLens AI - MCP Server

Connect PageLens AI reports to Claude Desktop, Cursor, Codex, and other MCP clients so your AI assistant can understand what failed, why it matters, and what to fix next.

An MCP (Model Context Protocol) server that gives AI agents direct access to PageLens AI - the independent launch reviewer for AI-built websites. PageLens AI turns launch risk across UX, SEO, Performance, Accessibility, Security, Conversion, and QA journey audits into plain-English priorities and agent-ready fix context.

Plug it into Cursor, Claude Desktop, Codex, or any MCP-compatible client and your agent can read scan results, drill into findings, surface quick wins, record owner decisions, and help close the fix loop without leaving the IDE.


MCP Endpoint

https://www.pagelensai.com/api/mcp

This is a remote MCP server — no local install required.


Which PageLens Path Is This?

  • Launch Pack is for founders, marketers, and AI builders who want a one-off repair loop: owner-first verdict, fix prompts for their AI builder, Markdown export, desktop + mobile review, and a re-scan.
  • CLI/API/deploy hooks are for Pro+ users who want PageLens AI to create scans from CI, release scripts, and client workflows.
  • MCP is for AI assistants that need to read a PageLens report, understand the owner launch verdict, inspect the evidence, and help you patch or triage findings.

If you do not want to read technical detail, start in the PageLens web report and copy the fix prompt into Lovable, Bolt, Replit, Cursor, Codex, Claude Code, Copilot, or Windsurf. If you do want your coding agent to work directly with the report data, connect this MCP server.


Quick Start

Cursor

Add to your ~/.cursor/mcp.json (or workspace .cursor/mcp.json):

{

"mcpServers": { "pagelens": { "url": "https://www.pagelensai.com/api/mcp" } } }

Claude Desktop

Add to claude_desktop_config.json:

{

"mcpServers": { "pagelens": { "type": "http", "url": "https://www.pagelensai.com/api/mcp" } } }

Codex CLI

Add to ~/.codex/config.toml:

[mcp_servers.pagelens]

url = "https://www.pagelensai.com/api/mcp"

Generic / other clients

Use the streamable HTTP endpoint:

https://www.pagelensai.com/api/mcp

Authentication

This server uses OAuth 2.0. Your MCP client will open a browser window to authorize with your PageLens account on first connect. No API keys to manage.

Scopes granted on authorization:

ScopeWhat it unlocks
read:scansList and read your scan results
read:findingsRead findings and quick wins
write:feedbackSubmit finding feedback and owner decisions

Available Tools

whoami

Confirm which PageLens account this MCP session is operating on, plus granted OAuth scopes.

{}

list_domains

List the domains you've verified ownership of, along with badge tier and the scan currently anchored to the public badge.

{

"include_unverified": false }


list_scans

List your most recent PageLens scans. Filter by status, domain, or date. Returns a slim summary per scan (id, URL, score, grade, severity counts, and launchVerdict). Scans may include launchContext with builderPlatform and launchMoment so agents can frame fixes for the owner's workflow without changing evidence or severity. QA Audit scans include a compact qaAudit summary with confidence, journey-step count, pages reviewed versus page budget, blocked/needs-review step counts, auth-profile status, and the reason fewer than the page budget were captured when the safe same-origin link graph is exhausted.

{

"limit": 20, "status": "COMPLETE", "domain": "example.com", "since": "2026-01-01T00:00:00Z" }

Status values: PENDING · RUNNING · COMPLETE · FAILED · CANCELLED


get_scan

Read the full summary for a single scan: score, grade, owner-first launchVerdict, severity counts, launch context, executive summary, top-5 highest-priority findings, and per-persona reviews. For QA Audit scans, get_scan also returns a qaAudit block containing the product-flow synthesis, journey replay, safe blocked actions, confidence, authenticated-route context, and page-budget coverage.

{

"scan_id": "clxxxxxxxxxxxxxxx" }


Available Resources

pagelensai://scan/{id}/markdown

Fetch the same agent-flavoured Markdown report available from the PageLens UI. Standard reports include launch-context and verdict front matter such as ai_builder, launch_moment, launch_verdict_status, and launch_verdict when captured, followed by an Owner launch verdict section and an AI-builder priority prompt pack your agent can act on before it reaches the detailed evidence. For QA Audit scans, this includes front matter such as qa_journey_event_count, qa_confidence, and qa_needs_review_step_count, followed by the application interpretation, journey replay, blocked/risky paths, safe actions, and next QA tests.

Use this when an agent needs rich context to reason about a QA Audit:

pagelensai://scan/clxxxxxxxxxxxxxxx/markdown
pagelensai://scan/{id}/summary.json

Fetch compact JSON for a scan. Standard scans include the same launchVerdict object returned by list_scans and get_scan: status, label, headline, summary, and next step. For QA Audit scans, the qaAudit object includes journey metadata, synthesis, page-budget coverage, and any queue_exhausted reason explaining why fewer than the purchased page count were discoverable.

pagelensai://scan/clxxxxxxxxxxxxxxx/summary.json

list_findings

Page through all findings for a scan. Filter by severity, category, persona, page URL, or rule ID. Use format: "full" to include descriptions, suggestions, and evidence.

{

"scan_id": "clxxxxxxxxxxxxxxx", "severity": "HIGH", "category": "SECURITY", "format": "full", "limit": 50 }

Severity levels: CRITICAL · HIGH · MEDIUM · LOW · INFO

Categories: UX · SEO · PERFORMANCE · ACCESSIBILITY · SECURITY · CONTENT · HEADERS · DESIGN · ERROR

Personas: MARKETER · CRO · UX · ACCESSIBILITY · BRAND · EXECUTIVE · PERFORMANCE · SEO


get_quick_wins

Return the top N quick-win findings — high impact, low-to-moderate effort — ranked by the same Impact × Effort scorer used in the PageLens dashboard. Optionally override the scan's preset to re-rank under a different lens.

{

"scan_id": "clxxxxxxxxxxxxxxx", "limit": 5, "preset_override": "CONVERSION" }

Presets: PRE_SALES · PRE_LAUNCH · CONVERSION · INVESTOR · BRAND_POLISH


report_finding_feedback

Flag a finding as a false positive, wrong severity, wrong category, or not actionable. Requires a paragraph of reasoning and a concrete evidence snippet.

{

"finding_id": "clxxxxxxxxxxxxxxx", "kind": "FALSE_POSITIVE", "reason": "The selector .pointer-events-auto matches 100+ utility classes, not a single offending element.", "evidence": " — Tailwind utility, not an event-handler.", "proposed_severity": "LOW" }

Feedback kinds: FALSE_POSITIVE · INCORRECT_SEVERITY · INCORRECT_CATEGORY · NOT_ACTIONABLE · OTHER


acknowledge_finding_decision

Attach owner-controlled context to a finding when the issue is real, but reflects an intentional architecture, security, or product tradeoff.

This does not hide the finding, edit the report, or change the PageLens score. It records the rationale so the report can show that the owner has acknowledged the tradeoff, and future scans can recognise the same finding as previously acknowledged.

{

"finding_id": "clxxxxxxxxxxxxxxx", "decision": "INTENTIONAL_TRADEOFF", "reason": "Next.js Cache Components and PPR currently prevent us from using per-request CSP nonces safely.", "evidence": "proxy.ts documents the CSP tradeoff: script-src uses 'unsafe-inline' because cached HTML shells cannot vary nonces per request.", "expires_at": "2026-10-01T00:00:00Z" }

Decision kinds: ACKNOWLEDGED · ACCEPTED_RISK · INTENTIONAL_TRADEOFF · WONT_FIX_NOW


clear_finding_decision

Clear a previously acknowledged decision so it stops appearing on current and future reports. The audit history is preserved.

{

"decision_id": "clxxxxxxxxxxxxxxx", "reason": "We have migrated to a nonce-compatible rendering path." }

You can also clear by finding_id when you do not have the decision_id.


What PageLens Checks

Each scan runs a deterministic rule engine + AI reviewer pipeline across every page:

AreaExamples
SEOTitle/meta length, canonical URL, Open Graph, heading hierarchy
PerformanceCore Web Vitals (LCP, CLS, INP), page weight, render-blocking resources, DOM size, third-party load
SecurityHTTP→HTTPS redirect, exposed .env/.git files, SSL expiry, XSS surfaces, source maps, exposed secrets in page source
AccessibilityFocus-visible CSS, reduced-motion support, ARIA patterns
UXHero hierarchy, mobile menu patterns, CTA structure
ContentPlaceholder text, stale copyright year, mixed-content references
QA AuditJourney replay, safe form/input exploration, blocked risky actions, app-flow synthesis, page-budget coverage

Findings include severity, effort estimate, copy-pasteable evidence, and a one-line fix suggestion.

QA Audit scans are different from standard technical scans: the primary artifact is the agentic journey report. PageLens will attempt to review the purchased page budget (for example, up to 10 pages on QA Audit) and should only return fewer pages when the safe same-origin link graph is exhausted. It can use validated auth profiles for scoped post-login routes, while still avoiding SSO, CAPTCHA, MFA, signup, payment, destructive changes, and other committing actions.


Example Agent Workflows

Fix with my AI builder:

"Read the latest PageLens Launch Pack report for example.com, list the top 3 owner-risk fixes, and turn each one into a patch plan I can apply in this repo."

Respect launch context:

"Open the latest PageLens report. If it was built with Codex or Cursor and the launch moment is Product Hunt, prioritise the fixes that make it safe to post publicly, then give me exact patches."

Pre-launch audit in Cursor:

"Use the latest PageLens scan for my staging site, list all CRITICAL and HIGH findings, and create GitHub issues for the top 5."

CRO review:

"Fetch the latest scan for example.com, get the CONVERSION quick wins, and summarise what to fix before the campaign launch."

Security sweep:

"List all SECURITY findings with severity HIGH or above from my last scan and show me the evidence for each."

Accepted architecture decision:

"For the CSP unsafe-inline finding, acknowledge this as an intentional Next.js/PPR tradeoff with the rationale from our security docs. Do not mark it as a false positive."

Post-fix validation:

"I re-ran PageLens from the CLI/API after fixing the findings. Compare the latest scan to the previous one and tell me what improved, regressed, or still needs work."

MCP reads and works with report data through OAuth. To create a new scan from automation, use the PageLens CLI, API, GitHub Action, or deploy hooks with a Pro+ plan; then ask your MCP-connected assistant to inspect the completed report.

QA Audit review:

"Fetch the latest QA Audit for example.com, read the markdown resource, and tell me which user journeys were verified, which actions were blocked safely, and whether PageLens reviewed the full page budget."


Pricing

PageLens AI suppor

只显示了开头部分;其余内容请在仓库中阅读