swahili-native-copy
Pass
Audited by Gen Agent Trust Hub on Oct 11, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is composed entirely of linguistic references and instructions for Swahili translation. It does not include any scripts, executable binaries, or system configuration changes.
- [INDIRECT_PROMPT_INJECTION]: The skill's primary function is to process external data (source briefs and English captions) which presents a theoretical injection surface. However, the skill is explicitly scoped to "Read and search only" and lacks the necessary capabilities (such as network access or file writing) to facilitate data exfiltration or system compromise, rendering the risk negligible.
Audit Metadata