writing-shape

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill instructs the agent to read arbitrary user-supplied files end-to-end and to "pull a fragment" and place or paraphrase it into the article (and to verify cited files), which means the LLM may be required to reproduce secret values verbatim from the input — creating an exfiltration risk unless explicit redaction rules are enforced.