The Agent Skills Directory

[DATA_EXFILTRATION]: The skill explicitly instructs the agent to read existing configuration profiles from ~/.claude/plugins/config/. Accessing hidden application directories in the user's home folder is a data exposure risk, as these files often contain sensitive user profiles, preferences, or session information.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process content from external files (e.g., law-practice-cn/CLAUDE.md, skills/*/SKILL.md) and user-supplied legal materials to guide the agent's behavior.
Ingestion points: The agent is directed to read domain files from the law-practice-cn directory and various skill-specific files within that path.
Boundary markers: The instructions do not define specific delimiters or warnings to isolate instructions from the data contained within these external files.
Capability inventory: The skill relies on tools for local file reading, web verification, and document rendering.
Sanitization: There is no evidence of validation or sanitization for the content loaded from the referenced external files or user materials.

chinese-legal-law-practice-cn