skill-lens

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill mandates scanning files and including original sentences/code snippets as evidence (e.g., "必须引用原句作为证据" and using inspect_skill.py output), which can force the LLM to reproduce any API keys or secrets found in those files verbatim, creating an exfiltration risk.