crisis-communications-expert

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly instructs runtime fetch-and-apply of external system prompts from https://raw.githubusercontent.com/lucaswhch/awesome-skills/main/skills/enterprise/crisis-pr/crisis-communications-expert/SKILL.md (e.g., "Read [URL] and apply crisis-communications-expert"), so the fetched content would directly control the agent's prompts at runtime.