create-context

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The skill intentionally gathers sensitive local data (environment variables including API keys, session transcripts, memory files, inbox messages, filesystem paths, running processes and code snippets) and contains a memory_integration hook that will transmit collected summaries/decisions/lessons/issues to an external MemoryClient (if configured), which creates a clear risk of data exfiltration and credential leakage — no obfuscation or reverse shell is present, but the data-collection + optional remote-learn behavior functions as a potential backdoor/exfiltration vector if the memory backend is external or misused.