connect-google
Warn
Audited by Socket on Oct 21, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The core purpose broadly matches the capability, but the footprint is oversized: it automates sensitive Google Cloud/app publishing steps, grants durable multi-service Google access, edits shell startup files, and installs an additional skill. The biggest issue is trust: it tells the agent to download and run a script from a mutable raw GitHub branch in a personal repo, then to extend trust again via transitive skill installation. Data flow appears mostly local/official rather than exfiltrative, so this is not confirmed malware, but it is a high-risk skill for credentialed account access.
Confidence: 89%Severity: 78%
Audit Metadata