catsu

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt includes an explicit per-request API key override (api_key="sk-different-key-for-this-request") and export examples with sk-... placeholders, which encourages embedding secret values verbatim in generated code/commands and thus poses a high exfiltration risk.