reproduce-frontend-bug
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill targets external issue trackers (GitHub, Jira, etc.) as data sources. It includes specific safety markers to treat this data as untrusted evidence and explicitly forbids the execution of commands found within these reports without validation.
- [DATA_EXFILTRATION]: The skill handles diagnostic logs and browser traces that could contain user credentials. It mandates the redaction of such sensitive information, ensuring that captured artifacts do not lead to data exposure.
Audit Metadata