What problem does it solve? Many firms and in-house legal teams lack a written AI usage policy, or rely on outdated versions that ignore recent state AI laws, the EU AI Act, and current bar guidance. This Skill produces a sourced first-draft policy for attorney review instead of an invented or finished policy. ## Core Features & Use Cases - Scope interview before drafting: Presents a checklist of ten policy sections (scope, permitted uses, approval, disclosure, data handling, training, incidents, enforcement, review cadence, glossary) so the user picks what the draft covers. - Jurisdiction-matched sourcing: Searches published model policies (ABA Formal Op. 512, state bar guidance, ILTA, CLOC, EU AI Act, SRA, PDPC, and others) based on the practice profile's regulatory footprint, and records every source in a cited Sources block. - Review-flagged draft output: Marks every threshold, named tool, and judgment call with [review] flags and open questions, plus a header, reviewer note, and adoption checklist signaling the document is not for distribution. - Use Case: A 30-lawyer firm asks for an AI usage policy; the Skill interviews the user on scope and audience, searches current model policies, and produces a draft with inline citations and decision points for the managing partner to resolve. ## Quick Start Ask the assistant to draft an AI usage policy for your firm or legal team, optionally specifying a scope such as firm-wide, legal team only, or an update to an existing policy.
