The Agent Skills Directory

[Local Configuration Management]: The skill reads and writes to specific local paths (e.g., ~/.claude/plugins/config/claude-for-legal/) to manage matter-specific context and handbook settings. This allows the agent to maintain consistency across different legal projects.
[Processing Untrusted Content]: The skill is designed to ingest and analyze policy changes provided by the user. While essential for its function, this represents a potential surface for indirect prompt injection if malicious instructions are embedded within the proposed handbook text.
[Command Execution]: The instructions reference the use of specific workspace commands (e.g., /employment-legal:matter-workspace) to switch context, which is a standard operational pattern for this type of legal-focused tooling.

handbook-updates