ad-director
Pass
Audited by Gen Agent Trust Hub on Mar 1, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill includes explicit safety rules in its instructions to refuse deceptive, illegal, or harmful advertising requests, demonstrating a security-conscious design.
- [SAFE]: Input handling in the
src/index.jsfile uses sanitization and request guarding functions (safeText,guardRequest) from a shared utility library to minimize the risk of malicious input processing. - [SAFE]: Validation scripts use controlled subprocess execution to run local Python scripts for checking output integrity, with no exposure to remote code or unauthorized commands.
- [SAFE]: The skill uses structured data envelopes for outputs, which helps maintain clear boundaries between generated content and agent instructions, reducing the risk of indirect prompt injection.
Audit Metadata