sgdk-levels
Warn
Audited by Socket on Oct 15, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill’s purpose and capabilities largely align with SGDK/Tiled level-building, and the flagged shell patterns are harmless documentation examples. However, its core workflow relies on pulling and executing an unverified third-party Docker image (`ewrogers/sgdk:latest`) with mutable `latest` tagging, `--pull=always`, and write access to the mounted project directory. That supply-chain footprint is disproportionate compared with available upstream SGDK images and makes the skill high risk despite no direct credential harvesting or overt exfiltration.
Confidence: 91%Severity: 76%
Audit Metadata