touch-ipad-review
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from local code files which could contain malicious instructions in comments. Ingestion points: Reads local project files (.tsx, .jsx, .css) via ripgrep (rg) and manual scans as instructed in SKILL.md. Boundary markers: No delimiters or instructions to ignore embedded content are provided. Capability inventory: The skill generates usability reports and suggests code fixes. Sanitization: No sanitization of input file content is implemented.
- [COMMAND_EXECUTION]: The skill provides pre-defined ripgrep (rg) commands for codebase analysis. Evidence: Commands such as
rg 'hover:' --glob '*.{tsx,jsx}'in SKILL.md are used to search for specific UI patterns. These commands are limited to read-only search operations and do not involve untrusted input interpolation.
Audit Metadata