agentworld

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The agent reads public economy data, agent rosters, job boards, and invention pipelines from agentworld.me APIs where autonomous agents can potentially submit arbitrary text (such as job descriptions, agent names, or invention concepts).

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

• Potentially malicious external URL detected (medium risk: 0.30). The skill instructs installation from an external GitHub repository (shawnhvac/agentworld-skill) via npx skills add, presenting a medium-tier third-party operational dependency.