humanizer

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill explicitly requires preserving facts, names, numbers, quotes, and other source content in the rewrite and to output the final rewritten text (including file in-place rewrites), so any secrets present in the input would need to be reproduced verbatim by the LLM.