🔒 安全工具 — Top 10 AI Agent 工具

渗透测试、漏洞扫描、代码审计、密钥管理——让 AI 帮你做安全自检。适合开发者日常安全左移、运维合规审查,也适合安全工程师加速日常重复工作。

共收录 112 个工具 · 显示评分最高 10

1

oraios/serena

精选Python

A powerful coding agent toolkit providing semantic retrieval and editing capabilities (MCP server & other integrations)

Stars
100
2

millionco/react-doctor

精选TypeScript

Your agent writes bad React. This catches it

Stars
100
3

0x4m4/hexstrike-ai

精选Python

HexStrike AI MCP Agents is an advanced MCP server that lets AI agents (Claude, GPT, Copilot, etc.) autonomously run 150+ cybersecurity tools for automated pentesting, vulnerability discovery, bug boun

Stars
100
pip3 install -r requirements.txt
4

Tencent/AI-Infra-Guard

精选Python

A full-stack AI Red Teaming platform securing AI ecosystems via OpenClaw Security Scan, Agent Scan, Skills Scan, MCP scan, AI Infra scan and LLM jailbreak evaluation.

Stars
100
5

iflytek/skillhub

精选Java

Self-hosted, open-source agent skill registry for enterprises. Publish & version skill packages, govern with RBAC and audit logs, deploy on-premise with Docker or Kubernetes.

Stars
100
npm install -g @astron-team/skillhub
6

sheeki03/tirith

精选Rust

Terminal security for developers and AI agents. Intercepts homograph URLs, pipe-to-shell, ANSI injection, obfuscated payloads, data exfiltration, and malicious AI skills/configs before they execute.

Stars
100
brew install sheeki03/tap/tirith
7

snyk/agent-scan

精选Python

Security scanner for AI agents, MCP servers and agent skills.

Stars
100
8

eracle/OpenOutreach

精选Python

Linkedin Automation Tool: Describe your product. Define your target market. The AI finds the leads for you.

Stars
100
docker run --pull always -it -p 5900:5900 -p 6080:6080 -v ~/.openoutreach/data:/app/data ghcr.io/eracle/openoutreach:latest
9

cisco-ai-defense/skill-scanner

精选Python

Security Scanner for Agent Skills

Stars
100
10

openakita/openakita

精选Python

An open-source AI assistant framework with skills and agent architecture

Stars
100
pip install openakita[all]

💡 如何选择合适的 安全工具 工具?

评分是怎么算的?+

Trove 的评分包含 4 个维度:文档质量(Doc)、代码质量(Code)、社区活跃度(Community)和近期更新(Activity),总分 25 分,并划分为精选(≥20)、已验证(≥12.5)、社区(<12.5)三档。

如何选择适合自己的工具?+

先看 tier(精选/已验证优先),再看 stars 和近期更新时间,最后看 install_hint 确认安装方式是否符合你的环境(npm/pip/go 等)。

这些工具支持哪些 AI 平台?+

大多数工具以 MCP 协议为标准,兼容 Claude、OpenClaw、Codex、Cursor 等主流 AI 平台。部分工具还提供 OpenAI Function / LangChain Tool 格式。

如何把自己的工具上榜?+

在 GitHub 发布你的工具,带上相关 topic(如 mcp-server、ai-agent-skills),Trove 的爬虫会自动收录并评分。也可以到 opentrove.ai/publish 主动提交。

安全工具 - 最佳 AI Agent 工具 Top 10 | Trove