re-verify-requirements

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The prompt requires reading source files and reporting/recording exact "Actual Value" items in the generated verification report, so if any secrets (API keys, tokens, passwords) are present in the code or config those values would be included verbatim in the LLM output, creating an exfiltration risk.