yapyak
Pass
Audited by Gen Agent Trust Hub on Nov 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill integrates with external LLM providers (such as Anthropic, OpenAI, and Google) to translate user-defined strings extracted from source code, creating an indirect prompt injection surface.
- Ingestion points: Translatable strings are extracted from calls to
t()andt.as()within various application files (e.g., .tsx, .vue, .svelte, and .astro files). - Boundary markers: The instructions do not define specific delimiters or sanitization protocols to prevent embedded instructions in source strings from influencing the translation model's behavior.
- Capability inventory: The skill utilizes network capabilities to transmit extracted content to third-party AI APIs during the translation process (e.g., via the
yapyak translateCLI command). - Sanitization: No explicit sanitization or validation of the translatable content is described prior to its transmission to external AI services.
Audit Metadata