agentbox-inference

Warn

Audited by Socket on Mar 18, 2026

1 alert found:

Security
SecurityMEDIUM
SKILL.md

SUSPICIOUS: The skill is coherent with its stated purpose, but it sends prompts through a third-party inference proxy and triggers automatic cryptocurrency micropayments for each call. The main risk is intermediary data handling plus autonomous real-world payment actions, not clear malware or hidden behavior.

Confidence: 85%Severity: 74%
Audit Metadata
Analyzed At
Mar 18, 2026, 05:09 PM
Package URL
pkg:socket/skills-sh/cascade-protocol%2Fagentbox%2Fagentbox-inference%2F@59b4038362bb447d463ca2c05918ca239e387668