caveman
Fail
Audited by Snyk on Jul 29, 2026
Risk Level: HIGH
Full Analysis
HIGH W007: Insecure credential handling detected in skill instructions.
- Insecure credential handling detected (high risk: 0.90). The prompt requires preserving code blocks and "exact error strings" verbatim, so any user-provided secrets/API keys pasted into input would be echoed unchanged by the LLM, creating a high exfiltration risk.
Issues (1)
W007
HIGHInsecure credential handling detected in skill instructions.
Audit Metadata