instrument

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). This guidance intentionally wires application code to send detailed telemetry (including agent configuration and runtime inputs) to an Opik endpoint, makes functions discoverable/remote-invocable via "entrypoint" metadata, and persists credentials/config in user files — behaviors that can be used to exfiltrate sensitive environment variables, API keys, or user data to an external service and therefore present a high risk if the target Opik endpoint is untrusted or malicious.