The Agent Skills Directory

[SAFE]: The skill performs legitimate legal analysis by organizing provided case materials into a deposition outline. It follows a structured workflow to categorize witness roles and relevant evidence.
[DATA_EXPOSURE]: The skill accesses a configuration file at ~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md. This path is specific to the application's local plugin configuration directory and is used to retrieve context necessary for the legal analysis, posing no risk to general system credentials.
[PROMPT_INJECTION]: The skill ingests external data ("case materials") to extract facts. This represents a standard surface for indirect prompt injection. However, the skill implements strict instructions for "Verbatim Quotes" and "Precise Citations," requiring the agent to verify every proposition against the record, which acts as a practical safeguard against following instructions embedded within the processed documents.

deposition-prep