meanest-editor
Fail
Audited by Snyk on Jun 16, 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 explicitly requires the model to "Always quote the draft directly" and reproduce user lines verbatim (line-by-line quoting), which would force the LLM to output any secret values present in a submitted draft, creating an exfiltration risk.
Issues (1)
W007
HIGHInsecure credential handling detected in skill instructions.
Audit Metadata