name: Privilege Sentinel
practice_area: Compliance / AI Governance
jurisdiction: Multi
version: "1.0"
author: Emily Cabrera — Hackathon 2026
description: >
  Pre-flight privilege and work-product check for legal AI prompts.
  Evaluates whether a prompt about to be sent to a third-party AI
  surface (ChatGPT, Claude.ai, Copilot, Gemini, etc.) risks blowing
  attorney-client privilege or work-product protection. Returns a
  SAFE / CAUTION / STOP band with cited factors, a discovery-impact
  line, and a redacted-safe rewrite. Backed by a knowledge base of
  case law, ethics rules, surface profiles, and a risk taxonomy.