gmgn-callout-score

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

• Third-party content exposure detected (medium risk: 0.30). The workflow reads caller-written text fields (call_thesis.source_content, token symbols, and profile name fields) from the GMGN API, which originate from external third-party users (callers and token creators).

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

• Potentially malicious external URL detected (high risk: 0.70). The skill instructs the agent to check IPv6 status via an external URL fetch (curl https://ipv6.icanhazip.com), presenting a weak/unknown/suspicious operational dependency during a runtime troubleshooting flow.