Deterministic pre-action authorization for AI agents. Guardrails run before tool execution, so prompt injection cannot bypass policy checks.
Deterministic pre-action authorization for AI agents. Guardrails run before tool execution, so prompt injection cannot bypass policy checks.
Maintain this project? Show that it is featured on YunoDB.
[](https://yunodb.com/tools/aporthqaport-agent-guardrails?utm_source=github&utm_medium=badge)<a href="https://yunodb.com/tools/aporthqaport-agent-guardrails?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/aporthqaport-agent-guardrails.svg" alt="Featured on YunoDB" height="20"></a>
FinceptTerminal is a modern finance application offering advanced market analytics, investment research, and economic data tools, designed for interactive explo
A 100% free modern JS SaaS boilerplate (React, NodeJS, Prisma). Full-featured: Auth (email, google, github, slack, MS), Email sending, Background jobs, Landing
🤖 Multi-platform IM AI Agent for Telegram, WhatsApp, Lark, and WeChat. Connects ChatGPT / Claude / Kimi / DeepSeek / Ollama / Pi for auto-replies, community an