blog-image
Warn
Audited by Socket on Oct 15, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The core direct-API workflow is coherent and mostly benign for a blog-image skill, using official Google docs/domains and proportionate local file writes. However, the optional MCP path materially increases risk by installing a third-party npm package and forwarding Google API credentials to it; this is not strictly necessary for the stated purpose and creates a meaningful trust boundary. No evidence of confirmed malware, covert exfiltration, or pre-execution payloads was found.
Confidence: 89%Severity: 68%
Audit Metadata