verification-before-completion

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). This skill requires pasting the exact commands run and the full captured stdout/stderr as evidence, which would force the agent to reproduce any secrets present in command lines or outputs (API keys, tokens, cookies, passwords) verbatim and thus creates an exfiltration risk.