brief-section-drafter
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a standard productivity tool for legal drafting and does not contain malicious code or deceptive instructions.
- [DATA_EXPOSURE_&_EXFILTRATION]: The instructions direct the agent to load a local configuration file at
~/.claude/plugins/config/claude-for-legal-zh/litigation-legal/CLAUDE.mdto ensure stylistic and theoretical consistency. This is a routine configuration practice for AI plugins and no exfiltration vectors (such as network calls) are present. - [INDIRECT_PROMPT_INJECTION]: The skill ingests user input related to legal cases. However, it lacks dangerous capabilities—such as file modification, network communication, or arbitrary command execution—precluding any significant risk from indirect prompt injection.
- Ingestion points: User-provided case details and section requests in
SKILL.mdprompt context. - Boundary markers: None explicitly present.
- Capability inventory: No tool invocations, shell commands, or network operations are requested or utilized.
- Sanitization: Not applicable given the lack of exploitable capabilities.
Audit Metadata