reviewing-english-prompts

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). This skill repeatedly requires preserving and returning the user's original prompt and verbatim tokens (e.g., "Keep exact product names, APIs, identifiers, quoted errors, and code tokens unchanged" and "send the exact user prompt to a reviewer"), which means any secrets embedded in the user's prompt would be echoed by the LLM and thus exposed.