sgdk-ai

Warn

Audited by Socket on Oct 16, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

SUSPICIOUS: the skill’s purpose and capabilities are coherent as an SGDK AI guide, and there is no credential access, exfiltration, stealth behavior, or dangerous pre-execution syntax. The main issue is install trust: it routes builds through an unverified third-party Docker image on a mutable `latest` tag rather than a clearly official, pinned SGDK image.

Confidence: 88%Severity: 72%
Audit Metadata
Analyzed At
Oct 16, 2026, 02:58 AM
Package URL
pkg:socket/skills-sh/ewrogers%2Fsgdk-skills%2Fsgdk-ai%2F@793909cea1d0641ef63841b6ec3318d1fcc87adfcd45f0fe1059edf0a84638af