Keeper Security MCP
Read-only Keeper MSP posture across every client, with history, a change feed and no path to a
作者提供的说明(英文原文)
Site/email/agent safety from public data: SPF/DKIM, SEO, monitor, hidden-AI, WHOIS, links (6).
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。
APIFY_TOKEN标为机密的值是你账户的钥匙。千万不要分享给别人,也不要粘贴到聊天里。
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 …”。
打开 Claude Desktop → Settings → Developer → Edit Config,把下面的内容粘贴到 claude_desktop_config.json
{
"mcpServers": {
"panda-web-security-tools": {
"command": "uvx",
"args": [
"panda-web-security-tools"
],
"env": {
"APIFY_TOKEN": "your-value"
}
}
}
}在终端中运行这条命令
claude mcp add panda-web-security-tools --env APIFY_TOKEN=your-value -- uvx panda-web-security-tools运行前请先填写空着的值。
粘贴到项目中的 .cursor/mcp.json,或 ~/.cursor/mcp.json
{
"mcpServers": {
"panda-web-security-tools": {
"command": "uvx",
"args": [
"panda-web-security-tools"
],
"env": {
"APIFY_TOKEN": "your-value"
}
}
}
}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": {
"panda-web-security-tools": {
"type": "stdio",
"command": "uvx",
"args": [
"panda-web-security-tools"
],
"env": {
"APIFY_TOKEN": "your-value"
}
}
}
}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。
这个服务器在你的电脑上运行,而 ChatGPT 只能连接互联网上的服务器。请选择其他应用。
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。
它在你的电脑上运行;安装后,工具列表会显示在你的 AI 应用中。
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10日
注册名称已通过 GitHub 账户 pandadad0214 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
它在你的电脑上运行,没有互联网端点
在你的电脑上运行,改用本地设置
它在你的电脑上运行;我们只能读取互联网端点的工具列表
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。
你需要填写:APIFY_TOKEN
它在你的电脑上运行。你的 AI 应用会在需要时启动它,数据不会经过作者的服务器(除非它自己调用在线服务)。
GitHub 仓库最后更新于–,最新版本(0.1.0)于今天进入注册中心。
支持 MCP 的应用,例如 Claude Desktop、Claude Code、Cursor 和 VS Code。你的电脑上还需要有 Python (uv)。
同一分类中有:Keeper Security MCP, BarzelVault: AI Action Authorization Firewall, pkgxray
同一分类中其他作者的服务器,按星标数排序。
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
Read-only Keeper MSP posture across every client, with history, a change feed and no path to a
Authorize AI agent tool calls before they run: policy-as-code, human approval, signed receipts.
Pre-install security scans for npm packages, MCP servers, and AI agents with cited verdict evidence.
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。
Give Claude, Cursor or any MCP client 6 site, email and agent-safety tools: SPF/DKIM/DMARC audits, AI search and SEO checks, page change monitoring, hidden prompt-injection scans, bulk domain expiry and broken links. Each tool runs one public-data Apify Actor on your own Apify account and is priced per result (the hidden AI instructions scanner is free). Part of the Panda Federation toolkit (36 tools in four focused servers).
uvx panda-web-security-tools
uvx, args ["panda-web-security-tools"], env APIFY_TOKEN (Cursor JSON below).summary / agentSummary line before widening.{"mcpServers": {"panda-web-security-tools": {"command": "uvx", "args": ["panda-web-security-tools"], "env": {"APIFY_TOKEN": "apify_api_YOUR_TOKEN"}}}}
| Field | Value |
|---|---|
| Purpose | Focused MCP server with 6 site, email and agent-safety tools: SPF/DKIM/DMARC audits, AI search and SEO checks, page change monitoring, hidden prompt-injection scans, bulk domain expiry and broken links. |
| Use when | A task needs a read-only audit of public DNS, web pages or links: email spoofing risk, AI/SEO readiness, page changes, hidden AI instructions, domain expiry or dead links. |
| Do not use when | You need penetration testing or logged-in scans (not offered), or the task is research (panda-research-tools), funding (panda-public-funding-tools) or leads (panda-leads-tools). |
| Install | uvx panda-web-security-tools (stdio, Python 3.11+) |
| Auth | APIFY_TOKEN env var; runs bill to that Apify account. |
| Tools | 6, listed under Tools |
| Outputs | JSON rows plus a one-line summary / agentSummary; Excel/CSV files on the Apify run. |
| Side effects and cost | Starts Apify runs; pay per result; empty runs are not charged per result; run_max_charge_usd caps each call. Read-only toward third-party sites. |
From the umbrella package (same tools):
uvx --from panda-federation-tools panda-web-security-tools
From this repo instead of PyPI:
uvx --from git+https://github.com/pandadad0214/panda-web-security-tools panda-web-security-tools
Cursor / Claude / VS Code: command uvx, args ["panda-web-security-tools"], env APIFY_TOKEN.
Official MCP Registry name for this stdio package: io.github.pandadad0214/panda-web-security-tools (server.json). That name is separate from the hosted research server io.github.pandadad0214/panda-research-mcp. Registry clients install the PyPI package with uvx panda-web-security-tools. The git command above stays the install path from this repo.
email_domain_security_checkerai_search_seo_auditwebsite_change_monitorhidden_ai_instructions_scannerbulk_domain_whois_expirybroken_link_checkerEach tool runs one Apify Actor on your account. Empty runs are not charged per result; every call is capped by run_max_charge_usd. Full schemas and prices: umbrella README.
panda-research-tools — Scholarly integrity and papers: citation checks, literature and PubMed search, academic experts, PDF tables, Wayback Machine snapshots.panda-public-funding-tools — US federal money, filings and safety monitors: grants, contracts, SBIR/STTR topics, SEC financials, clinical trials, Federal Register, recalls, VIN decode.panda-leads-tools — Public leads and enrichment: NPI providers, podcasts, local businesses, hiring signals, website enrichment, team pages, topic trends, trucking carriers, IRS nonprofits, building permits, new Florida, Texas and New York businesses, H-1B visa sponsors, EU tenders.Index / compatibility mega-server: panda-federation-tools (use only when you truly need every tool).
New tools join this domain only while the count stays in the 3–15 band. Descriptions follow the checklist in the umbrella repo. See tdqs-standard / project docs.
Built by Panda Federation. MIT licensed.