claude-peers-mcp-inter-instance-messaging
Warn
Audited by Socket on Sep 16, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The stated purpose is coherent for local peer messaging, but the skill asks users to disable Claude permission safeguards and enable development channels, which is disproportionate to the task. Install trust is also weaker because the skill is presented by ara.so while execution comes from a separate personal GitHub repo. No clear exfiltration or hidden behavior is shown, so this is not confirmed malware, but it is a high-risk skill to run as instructed.
Confidence: 88%Severity: 74%
Audit Metadata