repeat

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill explicitly logs the exact command and iteration outputs (including the "command" field and summaries) and prints iteration output, so any user-supplied command or prompt containing API keys, tokens, or passwords would be recorded and echoed verbatim, creating an exfiltration risk.