configuring-agent-brain

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The wizard and examples explicitly ask for and show API keys inserted verbatim into config files and export/cli commands (e.g., api_key: "sk-proj-..." and export OPENAI_API_KEY="sk-proj-..."), which requires the agent to accept and emit secret values directly.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs running remote install scripts that download and execute code (e.g., "curl -LsSf https://astral.sh/uv/install.sh | sh" and "curl -fsSL https://ollama.ai/install.sh | sh"), which fetch and run external code during setup and therefore pose a high-risk runtime dependency.