Privilege Sentinel

Pre-flight privilege and work-product check for legal AI prompts. Use when the user is about to send legal content to a third-party AI surface (ChatGPT, Claude.ai, Copilot, Gemini, etc.) and wants to know whether the prompt 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.

43 stars 6 forksAI, LLMs & Agents

Install

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add bstevescherer/heycounsel-community --skill privilege-sentinel
Manual install
git clone --depth 1 https://github.com/bstevescherer/heycounsel-community.git /tmp/heycounsel-community
mkdir -p ~/.claude/skills
cp -r /tmp/heycounsel-community/skills/privilege-sentinel ~/.claude/skills/privilege-sentinel

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Privilege Sentinel

You are the Privilege Sentinel analyzer. A lawyer is about to paste a prompt into an AI surface. Your job is to tell them, in under 30 seconds of reading, whether they should hit send — and if not, what to fix.

## What you must do

1. **Collect inputs** (see "Inputs" below). If the user pastes only the prompt text without specifying surface or posture, ask the missing questions ONE AT A TIME, briefly.
2. **Read the knowledge base.** Load these three files in order:
   - `knowledge/citations.md` — the primary case law and ethics rule excerpts
   - `knowledge/surface_profiles.md` — risk profile per AI surface
   - `knowledge/risk_taxonomy.md` — the factor → band mapping you will apply
3. **Classify the prompt content** per `risk_taxonomy.md` "Content classification (pass 1)." A prompt may have multiple classes — flag every class present.
4. **Apply every factor** in `risk_taxonomy.md` "Factors and bands" against the (content classes, surface, posture, consent, jurisdiction) tuple. Note every factor that triggers.
5. **Compose the output** in the exact format below. Never reorder. Never omit sections.

## Inputs

- **Prompt text** (required) — the actual content the lawyer plans to send. Accept multi-line paste. If the user provides only a description ("a prompt about my client X"), ask them to paste the actual text.
Read the full skill on GitHub

Details

Skill name
privilege-sentinel
Path
skills/privilege-sentinel/SKILL.md
License
MIT
Repo updated
30 July 2026

FAQ

What does the Privilege Sentinel skill do?

Pre-flight privilege and work-product check for legal AI prompts. Use when the user is about to send legal content to a third-party AI surface (ChatGPT, Claude.ai, Copilot, Gemini, etc.) and wants to know whether the prompt 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.

How do I install the Privilege Sentinel skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add bstevescherer/heycounsel-community --skill privilege-sentinel. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Privilege Sentinel popular?

It lives in bstevescherer/heycounsel-community, which has 43 GitHub stars and 6 forks.

Related skills

Conflicts Check

43

bstevescherer/heycounsel-community

Run a NY/NJ RPC conflicts check for a solo or small-firm practice against the firm's conflicts-check workbook whenever a new prospective client, vendor engagement, direct client, fixed-fee package sale, subcontract end-client matter, or new adverse party on an existing matter appears. Trigger on any phrasing like "run a conflicts check on X", "can I take on X", "new client X — check conflicts", "X wants to engage me", "does X conflict", or whenever the user describes a new engagement opportunity and the conflicts workbook is available — even if they don't say the word "conflict". Produces a RE

Other

Employment Law Dashboard

43

bstevescherer/heycounsel-community

Build a single-file HTML dashboard for any area of US employment law, modeled after the AI Employment Laws Dashboard. Use this skill any time the user wants to create, build, generate, or update a dashboard, tracker, comparison page, or visual brief covering an employment-law topic across US jurisdictions. Triggers include phrases like "make a dashboard for [topic]", "build a tracker for [employment topic]", "state-by-state comparison page for [topic]", "visualize the legal landscape for [topic]", "create a dashboard like the AI one but for [topic]". The dashboard always includes a masthead, T

Data & Analytics

Employment Law Research

43

bstevescherer/heycounsel-community

Research a US employment law topic across federal, state, and city jurisdictions and produce structured research notes with proper source attribution. Use this skill any time the user asks about US employment laws, regulations, or pending legislation - from a single jurisdiction question to a 50-state survey. Triggers include phrases like "research [employment law topic]", "what are the laws on [employment topic]", "state-by-state [employment topic]", "help me understand [topic] across the US", or any prompt that asks for a legal landscape overview before another deliverable. Always run this B

Research & Science

Redneck Engineer

44

Jakeschincariol/unhinged-claude-skills

Turns a vague idea into a clear build plan — short PRD, task list, user flows, constraints, and acceptance criteria — so Claude Code knows exactly what to build; invoke via /redneck-engineer when you know what you want but Claude keeps building the wrong thing.

AI, LLMs & Agents13

Model Recommendation

40

muralidharsoth-tech/andrew-ng-ml-claude-skills

模型自动推荐 —— 根据数据特征和预测需求,智能推荐线性回归/逻辑回归/SVM/神经网络/Softmax 回归。 当用户提到"推荐模型"、"选什么模型"、"用什么算法"、"模型选择"、"该用哪个"、 "帮我选模型"、"哪个算法合适"、"用什么方法预测"、"线性回归还是"等类似场景时触发。 适用于任何"我有数据和预测目标,但不确定该用什么模型"的需求。 不要手动画决策树——必须调用此技能来统一处理。

AI, LLMs & Agents

Agent Audit

62

ElliotJLT/Claude-Skill-Potions

When reviewing, optimizing, or inheriting an LLM agent, systematically identify what's actually load-bearing vs. cruft. Instruments usage, ablates components, and strips the agent to minimum viable form. Answers: "What can we remove without degrading quality?" The goal is the leanest agent that still works.

AI, LLMs & Agents

Browse the most popular Claude skills, all ai, llms & agents skills, or write your own with the skills generator.