design-audit
Pass
Audited by Gen Agent Trust Hub on Jun 26, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
Bashtool to perform local repository analysis and file searches using standard utilities such asgit,find, andgrep. These operations are limited to identifying UI-related files and searching for specific code patterns. - [PROMPT_INJECTION]: The skill ingests and processes content from untrusted source code files during audits, creating an indirect prompt injection surface (Category 8). Ingestion points: Local source code files (
.tsx,.jsx,.vue,.svelte,.css,.scss) read via thegreputility. Boundary markers: None identified in the prompt instructions. Capability inventory: Local command execution and file reading via theBashtool. Sanitization: No explicit content sanitization or instruction-ignoring delimiters are defined.
Audit Metadata