yes-md
Pass
Audited by Gen Agent Trust Hub on Mar 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides a comprehensive set of engineering guidelines and behavioral constraints designed to minimize errors and ensure accountability during task execution.
- [COMMAND_EXECUTION]: The instructions mandate the use of local system tools like
curl,cat,grep, andlsofto verify system state and validate changes. These operations are intended for local debugging and do not involve execution of untrusted remote scripts. - [DATA_EXFILTRATION]: Although the skill encourages inspecting diagnostic data (such as logs and environment variables), it does not contain instructions to send this information to external servers or unauthorized third parties.
- [PROMPT_INJECTION]: The skill uses structured instructional logic to impose safety gates and verification steps. While the language is authoritative, it serves to increase agent caution and does not attempt to override or bypass platform-level safety protocols.
Audit Metadata