“Is this allowed?” Assistant

About

Research AI Guardrails Assistant

An intentionally lightweight, local-first guide that applies policy, ethics, and privacy guardrails without relying on LLMs.

This MVP helps researchers and staff decide if a workflow is allowed, needs caution, or should be stopped. It is rule-driven and ships with a transparent library of guardrails and examples inside the repository.

  • Deterministic rules—no calls to external services.
  • Session-based storage only; no analytics.
  • Built with Next.js App Router, TypeScript, Tailwind, and shadcn-inspired components.

Disclaimer: This tool offers guidance, not legal advice. Confirm with your institution's policies and privacy/IRB officers.