tools-book-translation
Pass
Audited by Gen Agent Trust Hub on Feb 28, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill describes a legitimate workflow for translating documentation and UI components within a web project. No malicious patterns or adversarial instructions were detected.\n- [COMMAND_EXECUTION]: The skill uses basic shell commands (mkdir, cp) and project-specific scripts (node, npm) to facilitate the translation process. These actions are limited to the local file system and development environment, aligning with the skill's stated purpose.\n- [PROMPT_INJECTION]: The content includes instructional emphasis markers (e.g., 'IMPORTANT', '⚠️') used to guide the agent through necessary configuration steps. These are benign and do not attempt to bypass safety guidelines or reveal system prompts.\n- [DATA_EXFILTRATION]: No network calls to external or untrusted domains were found. The skill operates entirely on local project files within the established directory structure.\n- [CREDENTIALS_UNSAFE]: No hardcoded secrets, API keys, or sensitive environmental paths (e.g., .ssh, .aws) are accessed or exposed.\n- [SAFE]: [Indirect Prompt Injection Analysis] Ingestion points: MDX content files in src/content/book/ and messages/*.json. Boundary markers: Not explicitly defined. Capability inventory: Filesystem manipulation and execution of project-specific node/npm scripts. Sanitization: Not applicable as the content is trusted source material for translation. No vulnerabilities detected.
Audit Metadata