fable
Warn
Audited by Socket on Jul 20, 2026
1 alert found:
SecuritySecuritySKILL.md
MEDIUMSecurityMEDIUM
SKILL.md
SUSPICIOUS. The purpose is mostly aligned with the capability of routing work to Claude Fable 5, and official Anthropic model/CLI references are coherent. However, the skill's actual execution depends on unverifiable local scripts and a helper executable outside the reviewed text, so users must trust opaque code to probe availability, choose lanes, and spawn the model. That hidden executable footprint is disproportionate enough to classify the skill as suspicious rather than benign.
Confidence: 83%Severity: 78%
Audit Metadata