new-user-analysis
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill extensively uses the
ae-clitool to query business data, fetch project metadata, and verify authentication status. This is the primary mechanism for all analysis levels (L0-L4). - [DYNAMIC_EXECUTION]: To prevent shell escaping issues and ensure valid data formatting, the skill generates JSON analysis definitions dynamically using inline Python scripts before passing them to the CLI tool. This is a secure pattern for constructing complex tool inputs.
- [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it processes external data from ThinkingData event and user property tables. It mitigates this risk by requiring explicit user confirmation for analysis calibers, dimensions, and key behavior events before executing business data queries. It also uses structured JSON for all data exchanges with the CLI tool.
Audit Metadata