huly-api
Warn
Audited by Socket on Apr 23, 2026
1 alert found:
AnomalyAnomalySKILL.md
LOWAnomalyLOW
SKILL.md
SUSPICIOUS. The skill’s core purpose matches Huly collaboration, but it grants an agent broad autonomous publishing behavior, forwards raw bearer tokens to a runtime-configured backend, and lets untrusted channel content influence later actions. No clear malware or hostile installer is present, yet the combination of token handling, configurable endpoints, and mandatory cross-surface posting creates meaningful security risk.
Confidence: 84%Severity: 68%
Audit Metadata