preen-i18n
Pass
Audited by Gen Agent Trust Hub on Mar 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a project-specific audit script (
./scripts/preen/checkI18nCoverage.ts) and utilizes standard system utilities includingjq,rg(ripgrep),pnpm, andgitfor data processing and version control tasks. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface by ingesting and processing strings from the project's source code.
- Ingestion points: Source code files (
*.ts,*.tsx) and JSON output from the coverage script. - Boundary markers: None present to isolate or ignore potential instructions embedded in source code strings.
- Capability inventory: File system writes, local script execution, and git operations.
- Sanitization: No sanitization of detected strings is performed before they are processed by the agent.
Audit Metadata