Snapshot

io.github.Verifux由 GitHub 账号 Verifux 发布v1.0.4更新于 Nov 3, 2026

为任何网站免费提供UX和AI代理就绪度快照。无需注册,无需API密钥。

AI 翻译由 AI 模型根据下方的原文描述翻译,可能有误。

原文描述

Free UX and AI-agent readiness snapshot for any website. No sign-up, no API key.

已验证STDIO仅桌面AI 与机器学习

概览

Snapshot 的本地 MCP 服务;无需登录;STDIO。

暂无 AI 概览:以下信息来自注册中心条目和服务本身。

安装

在 SourceWeft 中

  1. 打开 控制台中的 Snapshot,将其添加到工作区。
  2. 为需要使用其工具的对话启用该服务。

Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。

其他 MCP 客户端

参照 仓库 中的启动说明。

README

Verifux snapshot MCP server

[npm] [license: MIT]

Can an AI agent actually use your website? This MCP server gives Claude, Cursor and any other MCP client one tool, run_snapshot, that audits any URL for UX, conversion and AI-agent readiness in a few seconds, and tells you whether the page even declares a surface an agent can act on.

Zero dependencies, no local browser, no API key, no sign-up. It calls the hosted Verifux API, so installing is just npx.

Install

Add it to your MCP client config:

json
{  "mcpServers": {    "verifux": {      "command": "npx",      "args": ["-y", "@verifux/mcp"]    }  }}

Then ask your agent: "run a Verifux snapshot on https://example.com".

What you get

run_snapshot({ url }) returns a plain-text report:

Verifux snapshot for your-store.comBand: Mixed - 13 of 17 automated checks passed0 critical, 2 major, 1 minor
  [MAJOR] Largest Contentful Paint at 3.1s, past the 2.5s threshold  [MINOR] Cumulative Layout Shift of 0.15, content moves under the reader
AI agent surface: none detected

That last line is the one most tools miss. As shopping and research agents become real traffic, a site that declares nothing an agent can call is invisible to them, and never sees why. Verifux is built to measure exactly that, alongside the human-facing UX.

Why it exists

Lighthouse and accessibility scanners tell you about code and compliance. Verifux asks a different question: can a person, and now an AI agent, actually get the job done on this page? The snapshot is the fast, free read. It is deterministic, so the same page scores the same way twice, and it carries no score out of 100 on purpose, because that number belongs to the full audit.

The full audit

This tool is the free snapshot. The full 54-checkpoint audit, across human experience (MX), behavioural design (BX) and AI-agent readiness (AIX), with a written report and a PDF, lives at verifux.app. No account, no subscription.

Notes

  • Point at a different backend with VERIFUX_API_URL (defaults to https://api.verifux.app).
  • The snapshot is email-optional through this client, so there is nothing to sign up for.
  • The hosted engine is closed source. This repository is the MCP client shim only: a small, zero-dependency wrapper that calls the public API.

Links

MIT licensed. Built by Designerama.

来源:README.md,提交 ac41b6b

工具

0
SourceWeft 尚未在沙箱中启动过该软件包,因此其工具未知。

版本历史

1
  1. v1.0.4最新Nov 3, 2026

评分

已满足适用检查项的 50%

该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。

  • 近 30 天内可连通25 分

    请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。

  • 已声明认证方式10 分

    请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。

  • 已列出工具15 分

    请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。

  • 描述必需15 分
  • 图标10 分
  • 文档15 分
  • 许可证10 分