What problem does it solve? Legal and compliance teams face constant informal requests like "can we just use AI for this?" without a consistent, documented way to answer. This Skill replaces hallway decisions with a fast, calibrated classification against your organization's own AI use case registry, red lines, and governance tiers. ## Core Features & Use Cases - Registry-based classification: Looks up proposed use cases in your configured registry and classifies them as APPROVED, CONDITIONAL, or NOT APPROVED, with reasoning and red-line checks. - Conditions and routing: For conditional approvals, produces a required-controls table with owners, governance tier, approval path, and a direct handoff into an AI impact assessment. - Cross-plugin and jurisdictional flags: Detects personal data (privacy/PIA handoff), product launches (product counsel handoff), and multi-jurisdiction conflicts such as EU AI Act or NYC LL144 exposure. - Batch triage: Processes a list or roadmap of use cases at once, outputting a summary table plus expanded detail for conditional and blocked items. - Use case example: A sales team asks to score leads with AI automatically — the Skill checks the registry, applies red lines, classifies the request, lists required conditions, and proposes a registry entry. ## Quick Start Ask the assistant to triage a proposed AI use case, for example: triage this use case — our sales team wants to score leads with AI automatically.
