localizing-software
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process external software resource files, such as string catalogs and ICU messages. These external files are untrusted and could contain malicious instructions (indirect prompt injection). The skill defines strict validation gates—including placeholder parity, schema checks, and syntax validation—which act as a secondary control to prevent the interpretation of data as commands.
Audit Metadata