veo3-1
Warn
Audited by Socket on Oct 18, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The code is coherent as a third-party video-generation client, and there is no clear malware, obfuscation, or hidden exfiltration beyond its stated API use. However, the skill's branding is materially misleading: it claims Google Veo 3.1 generation but actually sends prompts and bearer credentials to grsaiapi.com, an intermediary service rather than Google's official API. Combined with direct API-key CLI usage and an official-but-unpinned curl|sh installer for uv, this creates meaningful trust and data-flow risk even though the underlying behavior is straightforward.
Confidence: 94%Severity: 72%
Audit Metadata