The Agent Skills Directory

[Data Ingestion Surface]: The skill analyzes external documents such as PRDs, tickets, and marketing plans. While ingesting untrusted content can be a vector for indirect instructions, the skill incorporates specific safeguards, such as citation tiering and mandatory verification steps, to ensure outputs remain accurate and grounded.
[Workspace Configuration Access]: The instructions involve reading and writing to specific application paths within ~/.claude/plugins/config/claude-for-legal/. This behavior is used to manage framework calibration and matter-specific context, which is consistent with its intended purpose as a structured legal tool.
[Skill Chaining]: The process involves routing specific findings to other specialized components (e.g., privacy or AI governance skills). This modular design ensures that complex sub-tasks are handled by the appropriate logic while maintaining a clear audit trail within the legal workspace.

launch-review