The Agent Skills Directory

[SAFE]: The skill follows professional best practices for legal automation. It explicitly handles attorney-client privilege by providing instructions to separate privileged legal analysis from non-privileged project tracker updates. It also includes mandatory verification steps for AI-generated legal citations, preventing reliance on potentially fabricated or hallucinated references. The use of local configuration files for 'risk calibration' ensures the agent's behavior is grounded in the user's specific organizational context rather than generic model knowledge.

launch-review