code-analysis
Fail
Audited by Snyk on Apr 17, 2026
Risk Level: HIGH
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 1.00). The prompt's IRON LAW forces verbatim code snippets, exact trigger expressions, and example requests (e.g., curl/HTTP snippets) tied to cited file/line ranges, which would require the model to output any secrets present in source files or example commands, creating a high exfiltration risk.
Issues (1)
W007
HIGHInsecure credential handling detected in skill instructions.
Audit Metadata