business-logic-testing
Pass
Audited by Gen Agent Trust Hub on Jun 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructional content for security researchers and does not contain any prompt injection or safety bypass patterns.
- [SAFE]: No sensitive data exposure, hardcoded credentials, or unauthorized network exfiltration attempts were detected. The usage of placeholders like
$TOKENandURLfollows best practices for security documentation. - [SAFE]: The shell command example using
curlis a standard demonstration for testing race conditions in application logic and does not involve downloading or executing untrusted remote code. - [SAFE]: No obfuscation, hidden content (zero-width characters, homoglyphs), or metadata poisoning was found.
- [SAFE]: The skill describes an analytical process for evaluating application state machines and does not autonomously ingest or process untrusted data in a way that poses an indirect prompt injection risk.
Audit Metadata