Kakunin 是一个面向 AI 智能体的开源合规与身份平台,它通过 AWS KMS 签发 X.509 证书、实时评估行为风险、在风险超过阈值时自动吊销凭证,并生成符合 MiCA 和欧盟 AI 法案等法规要求的监管证据。

Stars

23

7 天增长

暂无数据

Fork 数

1

开放 Issue

1

开源协议

AGPL-3.0

最近更新

2026-07-09

AI 仓库情报摘要
FR-AI / ANALYSIS

为什么值得关注

它独特地将加密智能体身份(通过 AWS KMS 的 X.509 证书)与实时行为风险评分及自动凭证吊销结合起来,专为新兴 AI 监管需求设计,并以开源(AGPL-3.0)形式提供,同时保留托管信任锚以保证可验证性。

适合谁使用

  • 需要合规和身份基础设施的 AI 智能体开发者
  • 部署自主 AI 系统的组织中的合规与安全团队
  • 希望集成智能体监控和凭证管理的平台构建者
  • 评估智能体行为和审计追踪的监管机构或审计人员

典型使用场景

  • 当智能体行为超过风险阈值时自动吊销其凭证(例如智能体失控)
  • 为 AI 智能体操作生成符合欧盟 AI 法案或 MiCA 要求的监管报告
  • 为 AI 智能体签发和管理 X.509 证书以建立加密身份
  • 实时监控和评分智能体行为以检测异常或策略违规

项目优势

  • 实时行为风险评分直接关联凭证生命周期(自动吊销)
  • 使用 AWS KMS 进行密钥托管和 X.509 证书签发,满足企业安全标准
  • 开源平台(AGPL-3.0)配合 Apache-2.0 SDK,提供灵活的用法和集成方式
  • 从设计之初就致力于为 MiCA 和欧盟 AI 法案合规生成监管就绪证据

使用前须知

  • 自行托管平台需要配置大量外部服务(约 50 个环境变量),实施较为复杂
  • 信任锚(规范证书颁发机构)为托管服务,无法完全自主托管,限制了完全自治能力
  • AGPL-3.0 许可证可能要求运行修改版网络服务的用户提供源代码

README 快速开始

Kakunin — AI Agent Compliance Platform

An agent goes rogue → risk climbs → the certificate auto-revokes. Try it live, no signup: kakunin.ai/compliance-demo

Kakunin is compliance and identity infrastructure for AI agents. It issues X.509 certificates to agents via AWS KMS, scores their behavior in real time, revokes credentials automatically when risk crosses threshold, and produces regulator-ready compliance evidence — built for MiCA and the EU AI Act.

This repository is the platform / control plane: the Next.js application, the certificate authority integration, the behavioral risk engine, the API surface, and the compliance-reporting pipeline. It is the source behind the hosted service at kakunin.ai.

Open source, hosted trust anchor. The code is AGPL-3.0. The canonical certificate authority and public verification endpoint run as a hosted service, because a trust anchor everyone can fork is not a trust anchor — any counterparty must be able to verify a Kakunin certificate against one authority. See Open Source vs Hosted.

🚀 Most developers should start with the SDKs

You don't need to self-host this platform to use Kakunin. The fastest way to add cryptographic identity, behavioral monitoring, and compliance to your AI agents is the lightweight, Apache-2.0 SDKs — they talk to the hosted service, with free sandbox keys.

LanguagePackageInstallLicense
TypeScript@kakunin/sdknpm install @kakunin/sdkApache-2.0
Pythonkakuninpip install kakuninApache-2.0

TypeScript SDK → · Python SDK → · Docs →

Reach for this repo (Core) only when you need to run your own control plane. For everything else, the SDKs are the on-ramp.

What's here vs. the SDKs

The client libraries developers install live in their own repositories and are Apache-2.0:

RepoPackage
TypeScript SDKkakunin-sdk-typescript@kakunin/sdk
Python

项目描述

AI agent compliance platform — X.509 identity via AWS KMS, real-time behavioral risk scoring, auto-revocation, and MiCA / EU AI Act compliance reporting. AGPL-3.0.

相关仓库与替代方案

根据分类、Topic 和编程语言匹配的相似项目。

TanStack
精选
TanStack GitHub avatar

router

TanStack Router is a type-safe, data-driven React router with built-in caching, prefetching, and nested layouts, while TanStack Start extends it into a full-stack SSR framework.

Web 开发前端框架
14,861
vercel-labs
精选
vercel-labs GitHub avatar

scriptc

scriptc compiles ordinary TypeScript into small, fast native executables without needing Node.js, V8, or any JavaScript runtime in the binary.

开发者工具代码质量与构建
1,985
Jakubantalik
精选
Jakubantalik GitHub avatar

thinking-orbs

A React component library that renders six hand-tuned animated thought orb loading indicators on a plain 2D canvas, with two purpose-tuned sizes and automatic theme detection for AI and agent UIs.

AI 与机器学习AI 智能体
1,191