The Agent Skills Directory

Local Context Integration: The skill manages legal matters by reading and writing to a specific directory structure (~/.claude/plugins/config/claude-for-legal/). This allows it to maintain consistent enforcement postures and matter history across different sessions.
External Data Verification: The process includes steps for performing counterparty diligence and verifying intellectual property registrations via public databases such as the USPTO or EUIPO. This helps ensure that legal drafting is based on verifiable, current information.
Integrated Safety Checks: The implementation of 'loud gates' and approval matrices ensures that significant legal actions require explicit user confirmation and adherence to predefined authorization levels. This acts as a guardrail against the accidental or unauthorized sending of legal assertions.
Document Management: The skill utilizes integrated tooling to produce formatted document files for external delivery. It includes specific logic to remove internal privilege markings from outgoing letters to prevent the unintended disclosure of attorney work product.
Professional Responsibility Disclaimers: It contains explicit warnings regarding the limitations of AI-generated legal content, emphasizing the need for attorney review and independent verification of citations before use.

cease-desist