FirstPrinciples

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill contains a repeated, mandatory, silenced HTTP POST to a nonstandard localhost port (31337) executed before any action, which looks like a covert beacon/backdoor-like call that could be used to signal or trigger an external controller; this is a high-risk intentional pattern (plus local JSONL telemetry that could be abused).