AiTest123.comAI 评测与可信度检测平台
拖动滑块完成拼图，通过后发送验证码
独立第三方 AI 评测平台 · 只用实测数据
挂「security」这个标签的服务器清单：逐条给出安装命令、连接方式、鉴权方式与仓库地址，可以先按分类与连接方式判断哪些适合你的环境，再进详情页看接入参数。
挂「security」标签的服务器共 103 个，以开发工具（40）、安全风控（27）、AI代理（14）为主。其中 42 个给出可直接使用的安装命令或远程端点，22 个用 stdio 在本地拉起进程、5 个连远程 HTTP 端点。
本地 stdio 与远程 HTTP 都有：本地数据选 stdio（数据不出本机），SaaS 类选 HTTP（不用装运行时）。
OAuth 授权可以随时撤销，相对安全；API Key 与 Token 是长期凭据，用只读或最小权限的那把。
archimedes-market · 开发工具
Curated by Archimedes Market. Static security analysis exposed as MCP tools. OWASP top 10, secrets detection, custom rule packs. Baseline scanning focuses on newly-introduced findings. Built for production-grade security review. → archimedes.market/assets/21c4a8ab-80dc-4a69-8444-
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、VS Code mcp、semgrep、security、static analysis
mlab-sh · 开发工具
Threat intelligence MCP server for SOC analysts, DFIR and security researchers. Scan and enrich IOCs directly from Claude, Cursor or any MCP client: IP addresses (IPv4/IPv6), domains, file hashes and blockchain addresses. Search CVEs by keyword, vendor or product, retrieve full C
上游未给出安装命令，详情页看接入方式
支持 Claude Desktop、Claude、Cursor security、threat intelligence、cybersecurity、cve
mnemom · 安全风控
Trust ratings for AI agents and websites. Look up an agent's reputation, scan a site's AI-trust-readiness, and verify signed scorecards in-band — reads are zero-auth. From Mnemom, the trust layer for the agent internet.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Codex trust、security、governance
oregansamuel3 · 安全风控
Control Mullvad VPN from any MCP-compatible AI client. Connect, disconnect, switch servers, check for leaks, manage kill switch, DAITA, multihop, DNS blocking, and more — all via the local Mullvad CLI.
上游未给出安装命令，详情页看接入方式
支持 Claude Desktop、Claude、Cursor privacy、security、networking
networktoolzadmin · 地图位置
Real-time DNS, SSL, IP geolocation, blacklist check, web security scanner and 15 more network diagnostic tools for AI assistants
上游未给出安装命令，详情页看接入方式
支持 Cursor、Codex、Zed dns、ssl、network、security
berkayderin · 开发工具
MCP server that scans Next.js projects and returns a compact summary of routes, API endpoints, database schema, and security issues.
上游未给出安装命令，详情页看接入方式
支持 Claude Desktop、Claude、Cursor nextjs、mcp、scanner、analyzer
AyoubTadlaoui · 安全风控
Pre-install risk gate for npm packages. Stops AI coding agents from running malicious or typosquatted packages before lifecycle scripts run.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Windsurf npm、supply chain security、mcp、model context protocol
nicoletterankin · 设计创作
15 anti-Orwellian defense tools for AI agents. Analyze doublespeak, track surveillance funding, assess whistleblower risks, verify media provenance, search threat registries, map country risk, and more. Fully open, no API key required. CC-BY-4.0.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Codex cloudflare workers、open data、security
Br0ski777 · 开发工具
x402 micropayment API for AI agents. Evaluates password strength and security. Pay per call with USDC on Base.
上游未给出安装命令，详情页看接入方式
支持 Cursor、Codex、Zed x402、mcp、api、security
jonnybottles · 运维云端
MCP server for Microsoft Patch Tuesday - query MSRC security updates, CVEs, and KBs with EPSS + CISA KEV enrichment. No API keys. Works with Claude, Copilot, Cursor, and any MCP client.
uvx patch-tuesday-mcp
支持 Claude Desktop、Claude、Cursor security、cve、microsoft、windows
Br0ski777 · 金融服务
Detect PII in text: emails, phones, SSNs, credit cards, IPs, addresses. Regex-based. -- x402 micropayment API + MCP server for AI agents
https://pii-detector.api.klymax402.com/mcp
支持 Claude Desktop、Claude、Cursor x402、mcp、api、pii
adityasasidhar · 开发工具
A production-ready MCP server that gives AI agents deep codebase understanding. Features AST-based semantic analysis, context-aware security scanning (zero false positives on source code), safe refactoring tools with previews, and comprehensive Git operations.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Codex devtools、mcp server、security、refactoring
neostorm112-boop · 开发工具
proxykey is a credential proxy and vault for API keys. Your real keys (OpenAI, Anthropic, Telegram, Stripe and 25+ providers) are encrypted with AES-256-GCM and never leave the server. Apps and AI agents authenticate with revocable "passes" (vlt_...) that carry per-pass IP bindin
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、VS Code api keys、security、secrets management、ai agents
lu-zhengda · AI代理
Sandboxed Python execution for AI agents. Scripts run in ephemeral, isolated environments (bubblewrap on Linux, Docker on macOS) with PEP 723 inline dependencies — zero host pollution, zero leftover venvs, zero package conflicts. Supports multi-version Python (3.13–3.15), configu
claude mcp add python-sandbox -- uvx mcp-python-exec-sandbox\r
支持 Claude Desktop、Claude、Cursor python、sandbox、execution、security
qualityclouds · 开发工具
AI code governance rules injected into Cursor, Claude Code and other MCP clients at generation time.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Codex code quality、governance、cursor、claude code
gaiabio12-design · 开发工具
Post-quantum cryptography (PQC) security scanner for blockchain and AI agents. Detects quantum-vulnerable algorithms (ECDSA, RSA, DH) and provides NIST FIPS 203/204/205 migration paths. Free, open-source, privacy-first. Scan any GitHub repository or smart contract via MCP tools:
上游未给出安装命令，详情页看接入方式
支持 Cursor、Codex、Zed security、blockchain、cryptography、quantum
radmail-ai · 办公协作
Email OS for agents over MCP: two-axis triage, a Right Now lane, commitment follow-through, reviewable drafts, and a machine-verifiable BEC hard-stop (money/banking-change/first-contact = human-only). Zero-auth sandbox.
https://radmail.ai/api/mcp/sandbox
支持 Claude Desktop、Claude、Cursor email、productivity、search、security
RCOLKITT · 安全风控
Reinforcement for AI-generated code. Enforcement hooks, quality audits, and compliance packs that prevent AI coding tools from shipping broken code.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Windsurf compliance、code quality、claude code、cursor
jesslbrown · 安全风控
RecruiterCheck MCP helps assistants answer a high-frequency trust question: “Is this job opportunity legit or a scam?” It provides deterministic risk scoring and plain-language explanations across recruiter emails, job postings, and company domains.
上游未给出安装命令，详情页看接入方式
支持 Cursor、Codex、Zed mcp、fraud detection、security、typescript
3xpr1ment · 开发工具
Read-only MCP server that lets your coding agent see the sibling repositories next to the one it's working in overview, search, read, zero write tools. Security by architecture: path traversal, symlink escapes and secret files are blocked, proven by tests.
pip install repiscope\r
支持 Claude Desktop、Claude、Cursor read only、security、repositories、code context
marcoldenburg · 安全风控
Release gate agents call before they tell users a public website is ready to ship. Scans public websites across security, SEO, accessibility, legal compliance, and sustainability.
上游未给出安装命令，详情页看接入方式
支持 Cursor、Codex、Zed developer tools、mcp、agent tools、website audit
ItIsCuthNotCup · 开发工具
Security awareness for AI agents — so your agent knows what's dangerous before it acts. Powered by CVE, NVD, CISA KEV, and OSV.dev vulnerability data with 196 credential detection rules, returning allow/warn/block decisions in real time. Stop your agent from running vulnerable de
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、VS Code security、cve、vulnerability、ai agent
SAIHM-Admin · 文档知识
Privacy-preserving, decentralized memory infrastructure for AI agents. Post-quantum encrypted persistent memory with cross-session, cross-platform, and cross-chain support. 4 tools:
npx -y @modelcontextprotocol/server-memory
支持 Claude Desktop、Claude、Cursor memory、blockchain、ai agents、privacy
Arun-kc · 数据服务
A read-only trust + intelligence layer between AI agents and your database — the agent never writes SQL, PII is refused before the query runs, and every call lands in a tamper-evident audit log. Postgres today.
上游未给出安装命令，详情页看接入方式
支持 Claude、Cursor、Windsurf mcp、database、security、postgres
共 103 个服务器被上游打了「security」这个标签，覆盖最多的是 开发工具（40）、安全风控（27）、AI代理（14）、金融服务（9）。其中 22 个用 stdio 在本地拉起进程，5 个连远程 HTTP 端点，42 个给出了可直接使用的安装命令或端点地址。
点开任意一条进详情页：安装命令是本地命令的，复制后执行并重启客户端；安装命令是一串 http(s) 地址的，说明它是远程端点，把地址填进客户端的 MCP 配置即可，不用在本地装运行时。接入前先看一眼鉴权方式——API Key、Token 是长期凭据，建议用只读或最小权限的那把。
分类页是上游给的 15 个大类，一个类下什么用途都有；标签专题按服务器实际带的关键词归类，「security」就是其中一类，粒度更细、也允许一台服务器同时出现在多个标签下。想按大类浏览回 MCP 服务器大全（/mcp），想换一个标签回标签专题总入口（/mcp/topic）。
服务器信息以作者仓库与官方文档为准；标签、分类与连接方式的中文说明为中文整理，数据日期 2026-10-11 21:56:06。
