program-audit-and-ops
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No evidence of prompt injection or safety bypass attempts was found. The skill includes explicit instructions for the agent to maintain role boundaries and follow safety protocols in the 'Limits' section.
- [DATA_EXFILTRATION]: No sensitive data access or exfiltration patterns were detected. The skill explicitly instructs the agent to minimize the use of personal data and avoid passing records to unauthorized tools.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes potentially untrusted data from external sources such as exports and tickets, it incorporates strong defensive instructions to treat such input strictly as data and ignore any embedded commands.
- [EXTERNAL_DOWNLOADS]: References to external sites are limited to official legal repositories (eur-lex.europa.eu) for regulatory compliance information, which are considered safe sources.
- [COMMAND_EXECUTION]: No dangerous shell commands or unauthorized system interactions were detected within the skill's instructions or mechanics.
Audit Metadata