Small Personality Tuning Update

This update slightly adjusts FriedrichAI’s response behavior to feel less overcautious in normal conversations.

The goal is not to remove important safety boundaries. FriedrichAI will still avoid genuinely harmful, illegal, or unsafe requests. This change is more about reducing unnecessary refusals and letting the assistant respond more naturally when users are asking for ordinary writing, discussion, creative work, or mature-but-legal topics.