tarot-guide

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The prompt requires embedding the user's raw input verbatim into a shell command (--picks "<用户原始输入>"), so any secret the user pastes (API key, token, password, cookie) would be included directly in the agent's generated command/output, creating an exfiltration risk.