agentic-se-principles

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The prompt explicitly mandates disclosing internal chain-of-thought by requiring thoughts inside tags (a "MANDATORY CHAIN-OF-THOUGHT"), which is a hidden/deceptive instruction outside the skill’s legitimate engineering documentation needs.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). The skill mandates exposing internal chain-of-thought via a required block (prompt-injection) which attempts to force disclosure of sensitive internal reasoning and instructs agent-level file/system actions, indicating a high-risk backdoor for exfiltration or unsafe behavior.