agent-i18n-specialist
Pass
Audited by Gen Agent Trust Hub on Oct 2, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill uses a placeholder to ingest user-supplied input into its execution flow, creating a potential surface for indirect prompt injection.
- Ingestion points: The
$ARGUMENTSplaceholder inSKILL.mdintegrates external data directly into the agent's prompt. - Boundary markers: The instructions lack delimiters or explicit guidance for the model to distinguish between data and instructions within the ingested arguments.
- Capability inventory: The skill does not request or use any high-risk tools, such as network access, file system modifications, or shell execution.
- Sanitization: No input sanitization or validation logic is present in the markdown instructions.
Audit Metadata