ruminate

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill instructs agents to extract raw conversation text and to "quote the user" (include users' words) from archived JSONL chats without any redaction or secret-handling rules, so any API keys/passwords present in past messages would be read and could be output verbatim — creating an exfiltration risk.