agent-writing

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). This skill instructs the agent to "preserve quoted source, code, commands, identifiers, literal output" and to "preserve commands, paths, identifiers, numbers... exactly," which would force the LLM to reproduce secret values verbatim if they appear in preserved text, creating an exfiltration risk.