ez-google-imagen
Warn
Audited by Socket on Sep 16, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The stated purpose and Google credential scope are coherent, and the documented data flow points to official Google APIs rather than a proxy. However, the skill asks the agent to execute an unseen local script from ~/.openclaw/skills and may pull undeclared Python dependencies via uv, while also encouraging API key passing on the command line/chat; this is enough to keep risk above benign even without clear malicious behavior.
Confidence: 86%Severity: 52%
Audit Metadata