vietnam-visa-check
Pass
Audited by Gen Agent Trust Hub on Oct 19, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes a bundled Python script to process user queries. The script is self-contained, uses only standard libraries, and handles inputs through a safe normalization process, preventing arbitrary code execution.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface where it ingests untrusted user input via the nationality parameter. Ingestion points: User-provided country names or demonyms are passed to the query_visa.py script. Boundary markers: The agent is explicitly instructed to rely on the script's output fields and to avoid hallucinating. Capability inventory: The skill is restricted to reading its local JSON policy file and executing internal logic; it has no network or file-write capabilities. Sanitization: The input is sanitized using regex and matched against a hardcoded lookup table.
- [SAFE]: The skill is a legitimate utility with clear documentation. It uses bundled data to ensure privacy and reliability, and all referenced external links are to official government domains.
Audit Metadata