use-case-triage
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests and processes untrusted user data describing personal information processing activities to determine legal obligations.
- Ingestion points: User input provided via the
argument-hintand activity descriptions in the/use-case-triagecommand. - Boundary markers: The instructions lack explicit delimitation (e.g., XML tags or unique separators) to isolate the user-provided activity description from the logic of the prompt, potentially allowing a malicious description to influence the triage outcome.
- Capability inventory: The skill has read/write access to the local plugin configuration directory (
~/.claude/plugins/config/claude-for-legal-zh/privacy-legal/) and matter-specific files. - Sanitization: There is no explicit sanitization or validation of the input content beyond the agent's natural language understanding.
- [SAFE]: The skill implements a 'Destination Check' mechanism to prevent accidental data exposure to unauthorized parties or public channels.
- [SAFE]: The skill restricts file system operations to its own specific plugin configuration and project directories, following the principle of least privilege for its workspace.
Audit Metadata