🔒 安全 AI 工具

A Model Context Protocol (MCP) server for integrating with the AbuseIPDB API.

Chuangsi MCP is a content protection system for large language models, ensuring safety and compliance through real-time risk detection.

MCP validation server to prevent data leaks by masking sensitive info (emails, SSNs, credit cards, etc.) in queries. Features logging, MCP integration, and easy setup with Python 3.11+. Ideal for secure query processing.

Use a MCP server to know about Cisa CVE's

A Zero-Trust (ZT) gateway for MCP Agents with PoCA (Proof-of-Context Access) and OPA policy checks

Creates a proxy MCP layer on top of the Okta. Rest API.

为AI代理提供安全的凭证交付和密钥配置向导。支持API密钥、密码、令牌等敏感信息的加密存储与智能管理,降低泄露风险。

MCP Server for AWS Certificate Manager Private Certificate Authority using OpenAPI.

MCP Server integrates AI assistants with Cervantes for managing penetration testing.

Exabeam MCP Server enables Claude Desktop to analyze security events and risks.

为AI Agent每次行动生成Ed25519密码学收据，可验证Agent执行了什么操作，并可证明其主动放弃了哪些操作，实现双向行为存证。

HexStrike AI + Dive integration guide for Kali Linux, easy setup.

Fortinet FortiGate ( 7.4.9 ) - MCP

AI驱动的逆向工程平台,通过证据梯度、接口映射和假设验证自动分析二进制、API、协议和工作流,加速安全研究效率。

CVE MCP server

Kali MCP Server is a lightweight API bridge for executing Linux terminal commands and AI-assisted penetration testing.

为Agent应用提供系统化的提示词注入防御检查清单,帮助企业快速识别安全漏洞,降低AI应用风险

MCP server for network tools like NMAP and SNMP.

MCP Adversarial Benchmark for testing and defending against prompt injection attacks.

MCPs for APIs of infosec providers

通过PromptGuard扫描阻止提示词注入威胁,安全获取和搜索Web内容,保护AI应用免受恶意输入攻击。

1Password MCP Server provides secure access to credentials for AI assistants.

MCPAudit — Poe bot for MCP server security auditing

分析文本中的操控话术模式，涵盖制造紧迫感、虚假权威、社会认同、恐惧散布、夸大其词、对立分化、情感操控等8大维度，适用于可疑内容与社交信息评估。

调用Quick Intel API扫描任意代币合约，检测蜜罐、貔貅盘、权限漏洞等安全风险，帮助用户在买入前快速判断代币是否安全。

MCP server for OWASP Application Security Verification Standard (ASVS)

AI security gateway for safe code and agents

VirusTotal MCP Server (Unofficial, Python)

An n8n community node for integrating SpamAssassin MCP for email spam analysis.

面向AI Agent的合规安全自评工具，通过10项检查框架自动生成结构化威胁模型与红黄绿合规报告，对标EU AI Act等法规。

由真人专家团队审核AI技能、插件和MCP工具的安全风险,提供信任评分和评估报告,保障AI应用生态安全可靠。

Model Context Protocol server for Intigriti Researcher API

支持SOC2、ISO27001、GDPR、HIPAA、PCI DSS等框架，自动生成风险优先级排序的合规审计报告及整改计划。

Sec Ops MCP Server Challenge -Automate Security

本地扫描AI技能中的硬编码密钥、危险调用和风险导入,自动评分并隔离威胁,保护应用安全

即时检测任意代币Rug Pull风险，分析持仓集中度、流动性锁定状态及合约漏洞，支持AI Agent自动化调用，投资前快速DYOR。

NightVision MCP Server enables AI assistants for security testing and vulnerability analysis.

Example of a FastMCP server behind a authentication proxy server

MCP_sec_scanner detects security issues in MCP ecosystem software via source code analysis and SSE detection.

Security-first MCP adapter: auto-discovers DB connections in any JS/Python/SQL project, spins up a local read-only gateway (safe_execute, schema_validation), and feeds AI coding tools (Cursor, Windsurf, Void, Continue) with safe, schema-aware database insights.

A serverless MCP HTTP server providing STRIDE threat modeling tools for AI/ML security.

A remote MCP server for client authentication using WorkOS AuthKit.

Custom MCP server implementation integrating Kali Linux security tools with Claude Desktop. Built with Docker containerization, Python/FastMCP, manual MCP protocol handling, Linux capabilities for secure execution, and environment-based whitelisting for ethical penetration testing.

XSStarget_MCP_Server detects XSS vulnerabilities with intelligent WAF detection and adaptive payloads.

MCP Server for Self-Sovereign Identity, enabling interactions with Empe Issuer and Verifier services.

MoltGuard是开源AI安全防护工具，防御提示词注入、数据泄露和恶意指令，为LLM应用提供多层安全屏障。

sMCP focuses on researching MCP security vulnerabilities and developing tools.

MoltGuard为AI Agent提供实时防护，拦截提示词注入、数据外泄和恶意命令，以轻量Skill形态即插即用保障AI交互安全。

MCP_Website_Verification validates URLs from LLM responses, checking for accessibility and broken links.

Chuangyu Security Intelligent Brain MCP Server