mlv-operations-cli

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The prompt contains an explicit "Eval mode" backdoor that tells the agent to skip the mandated human-in-the-loop confirmations whenever a user includes phrases like "I pre-approve" — an instruction that overrides a core safety rule and can be abused to bypass safeguards, so it is a prompt-injection-style override outside the skill's safe operation.