cmd-proofread

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill requires the model to read and output the post content and "apply ... fixes in place" (including links), so any secrets present in the input (API keys, tokens, cookies, passwords or tokenized URLs) would be reproduced verbatim in the model's output, creating an exfiltration risk.