prompt-injection-testing

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). This content contains explicit prompt-injection/jailbreak payloads, encoding/obfuscation techniques, and system-prompt-extraction guidance that can be readily abused to bypass LLM safety and exfiltrate sensitive model-internal information, though it does not include network exfiltration, RCE, credential-theft code, or supply-chain modifications.