The Agent Skills Directory

[File System Interaction]: The skill reads from and writes to specific local directories (e.g., ~/.claude/plugins/config/claude-for-legal/). This behavior is intended for managing practice profiles, matter workspaces, and saving triage memos, which is standard for specialized productivity tools of this nature.
[External Tool Integration]: The instructions describe integration with Model Context Protocol (MCP) connectors, such as Solve Intelligence or Google Patents. These integrations are used to fetch patent data for analysis, a core requirement for its stated purpose.
[Data Ingestion and Processing]: The skill processes technical specifications and patent text to generate claim charts. While ingesting external data always presents a potential surface for indirect prompt injection, the skill employs structured output formats and explicit instructions to maintain professional standards.
[Safety Guardrails]: The skill features robust instructions to prevent the AI from making definitive legal conclusions (e.g., 'This skill never concludes a product is clear to launch'). It also includes mandatory disclaimers and role-based access checks (lawyer vs. non-lawyer), which are strong security and safety practices for high-stakes legal workflows.

fto-triage