gemini-translate
Fail
Audited by Snyk on Aug 18, 2026
Risk Level: HIGH
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 1.00). The skill instructs Claude to read full source files into prompts and explicitly to "preserve" credentials (i.e., not translate them) and then write translated files back, which forces the model to reproduce any secret values verbatim in its output.
Issues (1)
W007
HIGHInsecure credential handling detected in skill instructions.
Audit Metadata