Jitera-Labs/openguard

GitHub: Jitera-Labs/openguard

面向 LLM 应用的安全代理，拦截并清洗应用与 AI 服务商之间的流量，防止数据泄露和提示词注入攻击。

Stars: 0 | Forks: 0

![启动图片](https://static.pigsec.cn/wp-content/uploads/repos/2026/03/fe26d64454043810.png) ![Python](https://img.shields.io/badge/Python-3.10%2B-black?logo=python&logoColor=white) ![FastAPI](https://img.shields.io/badge/FastAPI-Framework-black?logo=fastapi&logoColor=white) ![Docker](https://img.shields.io/badge/Docker-Ready-black?logo=docker&logoColor=white) ![OpenAI Compatible](https://img.shields.io/badge/OpenAI-Compatible-black?logo=openai&logoColor=white) ![Anthropic Compatible](https://img.shields.io/badge/Anthropic-Compatible-black?logo=anthropic&logoColor=white) [![PyPI](https://img.shields.io/pypi/v/openguard?color=black&labelColor=black&logo=pypi&logoColor=white)](https://pypi.org/project/openguard/) [![Discord](https://img.shields.io/badge/Discord-Join%20chat-5865F2?logo=discord&logoColor=white)](https://discord.gg/7FSvyZZB) [![访客](https://api.visitorbadge.io/api/combined?path=https%3A%2F%2Fgithub.com%2FJitera-Labs%2Fopenguard&labelColor=%23000000&countColor=%23ffffff&style=flat)](https://visitorbadge.io/status?path=https%3A%2F%2Fgithub.com%2FJitera-Labs%2Fopenguard) OpenGuard 是一个 LLM 应用的安全代理。它位于你的应用程序和 AI 提供商之间，用于拦截、清洗和阻止流量。

标签：AI治理, AI防火墙, Anthropic兼容, API密钥检测, API网关, AV绕过, CISA项目, DLL注入, Docker, FastAPI, Linux系统监控, LLM应用防护, OpenAI兼容, Petitpotam, Python, 中间件, 人工智能安全, 合规性, 大语言模型安全, 安全代理, 安全防御评估, 护栏, 提示词注入防御, 数据脱敏, 无后门, 机密管理, 模型安全, 流量审计, 深度学习安全, 网络安全, 请求响应过滤, 请求拦截, 输入验证, 输出清洗, 逆向工具, 隐私保护