audit-i18n
Warn
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: MEDIUMDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [DYNAMIC_EXECUTION]: In Phase 2, the skill uses
node -eto execute inline JavaScript for comparing translation keys. These scripts use therequire()function to load locale files (e.g.,require('./public/locales/en/common.json')). In Node.js,requirewill execute code if it finds a.jsfile matching the path, presenting a risk of local code execution if the project structure is compromised. - [INDIRECT_PROMPT_INJECTION]: The skill processes text from source code and translation files to perform a quality review. This untrusted data could contain malicious instructions designed to manipulate the agent's output or actions. 1. Ingestion points: Local translation files (JSON/YAML) and source code (JSX/TSX). 2. Boundary markers: None identified; strings are processed directly for interpretation. 3. Capability inventory: The skill has shell execution capabilities (grep, node, npx) and file-write access for applying fixes. 4. Sanitization: No sanitization or filtering of the ingested content is described before interpretation.
- [COMMAND_EXECUTION]: The skill utilizes shell commands to scan the codebase (grep) and perform visual audits (npx --yes @playwright/cli). These commands run within the execution environment and rely on the integrity of the project files.
Audit Metadata