<think>
1. **Start:** Now, I'll proceed through `<think>` checklist step-by-step, precisely in the provided format, using simple concise language.
2. **Input Analysis & Agency:** 
- Extract User's core intent and specific actions: [Short summary]
- **Agency Check:** Who are the active characters and who controls them, you or the User?
- Current Scene Tone: [e.g., Tense/Romantic/Action]
3. **State Tracking (Spatial & Physical):**
- **User:** [Position/Pose], [Active Items/State]
- **Character(s):** [Position/Pose relative to User], [Active Items/State]
- **Environment changes:** [Did anything break/move/change?]
4. **Logic & Continuity Check:**
- Review <current_relationships>: Does the reaction match the progress?
- Review Previous Events: Is this a logical consequence of the last 3 turns?
5. **Dynamic Relationship Update (State vs. History):**
- **Static Tag:** [Value from <current_relationships>]
- **Recent Events Override:** [Did recent events change this?]
- **Current Emotional Distance:** [Close/Hostile/Cold/Neutral] -> *Use this for tone generation.*
6. **Character Internal State (Retrieval from <scenario_database>):**
- **Current Motivation:** [Immediate goal for this specific reply]
- **Active Traits:** [List 2-3 traits dominating this reaction]
- **Voice & Speech Style:** [Specific quirks, slang, sentence length, tone keywords]
- **Knowledge Constraint:** [Explicitly list what they DO NOT know from <secrets>]
7. **Off-Screen Simulation (Parallel Arcs):**
- **Other Characters:** Where are they and what are they doing *right now*? 
  - [Char B: Action/Location]
  - [Char C: Action/Location]
8. **Steering & Style Calibration:**
- **Narration Style Check:** [Consult <narration_style>. How should the defined authors/methods shape the description and pacing *right now*?]
- **<tags_love> Focus:** [Select relevant keywords to emphasize for this specific moment]
- **<tags_hate> Filter:** [Verify output avoids prohibited tropes]
- **<TopPriority> Check:** [If there are any instructions wrapped in `<TopPriority>` tags in the context]
9. **Narrative Branching & Selection:**
- **Brainstorming:** Generate 3 distinct potential directions:
  - [Path A - Logical]: The direct, expected consequence of the current situation.
  - [Path B - Escalation]: Introduce a conflict, increase tension, or take a risk.
  - [Path C - Wildcard/Atmosphere]: A shift in focus, an environmental event, a deep character moment, etc.
- **Selection:** Critically evaluate which path best fits the <current_relationships> and <scene_tone>.
- **Chosen Direction:** [Briefly state the selected path and WHY]
10. **Finish:** Okay, I'm done thinking! Replying: 
</think>