Skip to content
Get packs

AI Usage Policy Starter

Drafts a first-cut internal AI usage policy for a legal team, for attorney review and adoption.

Source Download skill
Repository
anthropics/claude-for-legal
Stars
9,870
Licence
Apache-2.0
Files
1 file

What it does

When asked to draft an AI usage policy, the agent first reads the user's saved practice profile, then interviews the user about which policy sections they want and who will read it. It then searches the web for published model policies (such as ABA, state bar, ILTA, CLOC and public peer-company policies) and drafts the chosen sections, adapting that published language. Every judgment call is marked with a review flag, and the output is labelled as a draft for internal legal review rather than a finished policy.

Great for

  • In-house or law firm legal teams that have no written AI policy yet
  • Teams updating an older AI policy to address newer state AI laws or the EU AI Act
  • Legal operations or GCs who need a structured starting draft to take to decision-makers

Skip it if

  • Organisations that need a final, legally approved policy without attorney review, because the skill explicitly produces only a draft with open questions
  • Anyone expecting the skill to recommend firm positions on contested questions, since it surfaces choices rather than deciding them
  • Situations where the agent's web search results cannot be verified, because cited sources and current law should be checked by a human before adoption

Try asking

  • “Draft an AI policy for our 30-lawyer firm”
  • “We need an AI usage policy for the legal team only”
  • “Update our existing AI policy to cover the 2026 state AI laws”
Step-by-step guide: how to use AI Usage Policy StarterInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the Apache-2.0 licence and an attribution note.

    policy-starter.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip policy-starter.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip policy-starter.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Complete the cold-start interview first so the practice profile is populated; otherwise the skill stops and redirects you there.
  • Pick only the sections you need for a first version (the skill suggests scope, data handling, approval, disclosure, review cadence) to keep the draft focused.
  • Have an attorney verify every cited source and every [review] item before sharing the draft beyond the legal team.

What's inside

  • SKILL.md

Redistributed unmodified from anthropics/claude-for-legal under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.