persian
Pass
Audited by Gen Agent Trust Hub on Mar 11, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user text via the {source_text} variable in SKILL.md. It uses boundary markers like 'SOURCE TEXT:' to delimit user content. Its capability inventory includes spawning six sub-agents with 'TeamCreate' and 'Task'. Sanitization is achieved through specific review mandates and a synthesis phase that prioritizes semantic fidelity.
- [DATA_EXPOSURE]: The skill references local files (TERMS.csv, PersianTerms.txt) containing linguistic and terminology data. These files contain domain-specific translation resources related to the Baha'i World Centre.
- [COMMAND_EXECUTION]: The orchestration of sub-agents is performed using internal environment tools (TeamCreate, Task) to manage the multi-agent team. These tools are used for agent orchestration and do not involve system-level shell access.
Audit Metadata