skills.rest
anthropics
anthropicsOFFICIAL
Explore all skills by this author →
WebsiteGitHub

policy-starter

Draft sectioned AI usage policy text with sources and review flags.

9.1k|1.8k|Updated Apr 21, 2026
Search Tags:#risk management#legal compliance#generative ai#policy drafting#ai governance#attorney review#sources and attribution
One-click install
npx skills add https://github.com/anthropics/claude-for-legal --skill policy-starter-anthropics
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: policy-starter
Source: https://github.com/anthropics/claude-for-legal/tree/main/ai-governance-legal/skills/policy-starter
Command: npx skills add https://github.com/anthropics/claude-for-legal --skill policy-starter-anthropics

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

It helps legal teams quickly create a first-cut internal AI usage policy that’s grounded in published model policies and ready for attorney review instead of starting from a blank page or drafting in isolation.

Core Features & Use Cases

  • Scope-first policy drafting: runs a guided scope interview to determine audience, tool coverage, and deployment context before writing any policy sections.
  • Model-policy sourced sections: drafts policy language section-by-section using relevant, jurisdiction-appropriate published guidance (with explicit reviewer flags at every decision point).
  • Attorney-adoption readiness: produces a structured draft with sources, in-line attribution, a reviewer note, and an adoption checklist so stakeholders can finalize safely.

Use case examples:

  • A 30-lawyer firm needs a new GenAI policy aligned to its jurisdictions.
  • A legal team updates an existing AI policy to reflect new regulatory expectations for AI governance.

Quick Start

Ask for a firm AI usage policy draft for internal legal review by running: /ai-governance-legal:policy-starter "we need an AI policy for our 30-lawyer firm".

Frequently Asked Questions about policy-starter

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I draft an AI usage policy for a law firm?▼

To draft an AI usage policy for a law firm, you can use a guided scope interview to determine tool coverage and deployment context, then generate sectioned, source-attributed policy text ready for attorney review.

What is included in a generative AI governance policy draft?▼

A generative AI governance policy draft includes sectioned text based on published model policies, explicit reviewer flags at decision points, in-line source attribution, and an adoption checklist for stakeholders.

Can I update an existing legal AI policy to reflect new regulatory expectations?▼

Yes, you can update an existing legal AI policy by loading your current matter context and performing model-policy web research to align the draft with new jurisdictional and AI governance regulatory expectations.

Does drafting an AI compliance policy require a blank page or prior context?▼

Drafting an AI compliance policy does not require a blank page; it uses a practice profile CLAUDE.md and optionally loads matter context to ground the initial draft in published guidance before attorney review.

How do attorney review flags work in AI policy drafting?▼

Attorney review flags in AI policy drafting work by inserting explicit reviewer notes at every decision point in the sectioned text, ensuring legal teams can safely finalize and adopt the generated policy.

Related Skills

Explore More Legal & Compliance Skills→
az9713az9713

policy-starter

Draft an internal AI usage policy with scoped sections and reviewer flags.

Community
Intermediate
uk-agentsuk-agents
9

policy-starter

Draft a UK internal AI usage policy from published model guidance.

Official
Advanced
jrhueiuengjrhueiueng

policy-starter

Draft a sourced AI usage policy with attorney-review flags.

Community
Advanced
PolliticalSolutionsPolliticalSolutions

policy-starter

Draft attorney-review AI usage policies with cited model-policy sources.

Community
Advanced