code-review

Fail

Audited by Snyk on Mar 6, 2026

Risk Level: HIGH
Full Analysis

HIGH W007: Insecure credential handling detected in skill instructions.

  • Insecure credential handling detected (high risk: 1.00). It mandates quoting exact lines from changed files as evidence in findings, which will force the model to reproduce any embedded API keys, tokens, or passwords verbatim in its output, creating an exfiltration risk.
Audit Metadata
Risk Level
HIGH
Analyzed
Mar 6, 2026, 04:33 PM