protocol-participation
Warn
Audited by Socket on May 11, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The skill is purpose-aligned for network labs and routing operations, and its dependency/install footprint appears mostly standard. However, it grants an AI agent direct authority to make live BGP/OSPF changes with real operational consequences, while the stated ServiceNow/audit guardrails appear procedural rather than technically enforced in the provided skill text. No strong credential-harvesting or exfiltration signal is visible, but the autonomy and network-impact risk are high enough to treat it as medium-high security risk.
Confidence: 86%Severity: 72%
Audit Metadata