intelligent-decision-making

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.80). The skill tells the agent to "Provide full context
• Include relevant files" and explicitly passes files=["src/auth/"] to the consult-llm MCP and logs decisions, which encourages sending authentication files (and any embedded secrets) to models and potentially outputting them verbatim.