The Agent Skills Directory

[SAFE]: The skill implements a professional legal workflow with a primary focus on patent claim analysis. It includes prominent disclaimers regarding the legal nature of its output and mandates review by registered patent counsel.
[SAFE]: The skill uses localized file system paths (~/.claude/plugins/config/claude-for-legal/) to store configuration and output files. These operations are scoped to its functional purpose and do not show signs of data exfiltration or unauthorized access to sensitive user data.
[SAFE]: External data sources are restricted to trusted platform integrations (MCPs) for patent research. The processing of external patent text is conducted within a structured claim-charting framework that mitigates typical prompt injection risks through technical formatting requirements.

fto-triage