The Agent Skills Directory

[SAFE]: The skill follows a well-defined instructional workflow for legal preliminary screening without any evidence of malicious intent or dangerous capabilities.
[SAFE]: The skill manages local state and configuration within its own specific directory (~/.claude/plugins/config/claude-for-legal/ip-legal/). This is standard practice for maintaining persistence and context in specialized agent plugins and does not involve unauthorized data access.
[SAFE]: The skill processes external data such as case facts and product descriptions. It includes clear guardrails ensuring that the AI does not provide final legal determinations, effectively managing the risk of indirect prompt injection.

infringement-triage