create-pr
Warn
Audited by Socket on Aug 20, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS: The skill’s Git/GitHub access and local inspection are broadly consistent with creating a PR, and no malicious download/exfiltration path is visible. However, it enables unconditional autonomous handoff and auto-merge without user confirmation, and part of the effective behavior depends on an unseen sibling skill/script, so the overall risk is medium-to-high despite low evidence of malware.
Confidence: 87%Severity: 76%
Audit Metadata