fix: protect server config from truncation and recover from Discord guilds

- Save servers_config.json atomically via temp file + fsync + rename
- Keep .bak backup and auto-restore when main config is empty/corrupt
- Add /servers/recover endpoint for manual recovery
- Auto-recover basic server configs on startup when config is empty but bot is in guilds

AGENTS.md

@@ -0,0 +1,82 @@
+# AGENTS.md
+
+## Language & runtime
+- **Python 3.11** (main bot). There is no root `package.json` or TypeScript — do not apply Node/TS tooling.
+- `uno-online/` is a secondary Node.js project; `miku-app/` is Android/Kotlin. Both shelved features for now.
+
+## Commands
+
+```bash
+# Build and run all core services (bot, STT, llama-swap, Cheshire Cat, Qdrant)
+docker compose up -d
+
+# Run with face-detector (requires NVIDIA GPU)
+docker compose --profile tools up -d
+
+# Run only the bot (implies dependencies are already up)
+docker compose up -d miku-bot
+
+# View bot logs
+docker compose logs -f miku-bot
+
+# Rebuild bot after code changes
+docker compose down miku-bot && docker compose build miku-bot && docker compose up -d miku-bot
+
+```
+
+## Config
+- **`config.yaml`**: app settings (model names, URLs, ports, feature flags).
+- **`.env`**: secrets only (`DISCORD_BOT_TOKEN`, `OWNER_USER_ID`, `ERROR_WEBHOOK_URL`).
+- Config is loaded by `bot/config.py` (Pydantic) and `bot/globals.py` (bare `os.getenv`). Both sources matter — check both when tracing config usage.
+- Runtime config overrides are persisted to `bot/memory/config_runtime.yaml` via the API.
+
+## Architecture
+
+```
+Discord <-> bot/bot.py (discord.py)
+               ├── on_message -> Cheshire Cat pipeline -> memory-augmented LLM response
+               ├── utils/llm.py -> llama-swap (HTTP proxy) -> llama.cpp (NVIDIA or AMD GPU)
+               ├── utils/voice_manager.py -> STT WebSocket (port 8766) and audio playback
+               ├── FastAPI (port 3939, daemon thread) -> 22 route modules in bot/routes/
+               ├── APScheduler (background tasks in globals.py)
+               └── utils/autonomous_engine.py -> proactive message decisions (Autonomous V2)
+```
+
+- The FastAPI server runs in a **daemon thread** inside the Discord bot process — no separate process.
+- `bot/globals.py` holds mutable global state (`scheduler`, env vars, `discord.Client`). Module-level mutations are pervasive; be careful with import order.
+- llama-swap is a llama.cpp HTTP proxy with TTL-based model swapping. Two configs: `llama-swap-config.yaml` (NVIDIA) and `llama-swap-rocm-config.yaml` (AMD).
+
+## Models (via llama-swap)
+| Model key | Purpose |
+|-----------|---------|
+| `llama3.1` | Primary text model |
+| `darkidol` | Uncensored model (evil mode) |
+| `vision` | MiniCPM-V (image understanding) |
+| `swallow` | Japanese text model |
+| `rocinante` | 12B model (AMD GPU only) |
+| `qwen3.5` | ComfyUI prompt generation (AMD GPU only) |
+
+## Testing & linting
+- **No formal test framework** and **no linting/formatting config**. Ad-hoc scripts live in `tests/` and `bot/tests/`.
+- Run ad-hoc tests however you want; there is no standard command.
+
+## Web UI color scheme (bot/static/)
+- **Base**: `#121212` body, `#000` log panel, `#1e1e1e` code blocks, `#2a2a2a` cards
+- **Text**: `#fff` primary, `#ccc` labels, `#888` muted, `#0f0` log info
+- **Primary accent**: `#61dafb` (headings, links, assistant messages, active elements)
+- **Success**: `#4CAF50` (active tabs, user messages, enabled toggles)
+- **Error**: `#f44336` (chat errors), `#ff6b6b` (error logs)
+- **Warning**: `#ffd93d` (warning logs)
+- **Bot message**: `#2196F3` (left border)
+- **Danger/evil**: `#ff4444` (overrides all accents when `body.evil-mode` is set)
+- **Bipolar**: `#9932CC` (toggle active)
+- **Blocked**: `#ff9800` (blocked user cards)
+- Evil mode toggles `body.evil-mode` class which replaces all `#61dafb` and `#4CAF50` with `#ff4444`.
+
+## Key gotchas
+- `bot/memory/` contains persisted JSON state files and is **gitignored**. Do not expect these to exist in a fresh clone.
+- `.env` is gitignored; copy `.env.example` to `.env` and fill in real tokens.
+- Changes to `bot/moods/` or `bot/persona/` text files take effect at runtime (loaded on demand), no rebuild needed.
+- Playwright browsers must be installed in the Docker image (`bot/Dockerfile` does this via `setup_uno_playwright.sh`).
+- Voice features require `discord-ext-voice-recv` and `PyNaCl` — if voice fails, check these are installed.
+- The `miku-voice` Docker network is declared as **external** — it must exist before `docker compose up`.

bot/Dockerfile

@@ -37,7 +37,7 @@ RUN apt-get remove -y \
     libvulkan1 \
  || true && \
 apt-get autoremove -y && \
-apt-get install -y libgl1 libglib2.0-0 && \
+apt-get install -y libgl1 libglib2.0-0 ffmpeg && \
 apt-get clean && \
 rm -rf /var/lib/apt/lists/*
 

bot/api.py

@@ -102,6 +102,7 @@ from routes.logging_config import router as logging_config_router
 from routes.voice import router as voice_router
 from routes.memory import router as memory_router
 from routes.activities import router as activities_router
+from routes.models_selector import router as models_selector_router
 
 app.include_router(core_router)
 app.include_router(mood_router)
@@ -123,6 +124,7 @@ app.include_router(logging_config_router)
 app.include_router(voice_router)
 app.include_router(memory_router)
 app.include_router(activities_router)
+app.include_router(models_selector_router)
 
 
 

bot/bot.py

@@ -163,6 +163,42 @@ async def on_ready():
     # Start server-specific schedulers (includes DM mood rotation)
     server_manager.start_all_schedulers(globals.client)
     
+    # Auto-recover server config if it was lost/corrupted (e.g., disk full)
+    if not server_manager.servers and globals.client.guilds:
+        logger.warning("⚠️ Server config is empty but bot is in guilds — attempting auto-recovery")
+        recovered = 0
+        for guild in globals.client.guilds:
+            text_channels = [ch for ch in guild.text_channels if ch.permissions_for(guild.me).send_messages]
+            if not text_channels:
+                text_channels = guild.text_channels
+            if not text_channels:
+                continue
+            preferred = None
+            for ch in text_channels:
+                if ch.name.lower() in ("general", "chat", "main", "lounge", "general-chat"):
+                    preferred = ch
+                    break
+            channel = preferred or text_channels[0]
+            try:
+                server_manager.add_server(
+                    guild_id=guild.id,
+                    guild_name=guild.name,
+                    autonomous_channel_id=channel.id,
+                    autonomous_channel_name=f"#{channel.name}",
+                    bedtime_channel_ids=[channel.id],
+                    enabled_features={"autonomous", "bedtime", "monday_video"}
+                )
+                recovered += 1
+                logger.info(f"🔄 Auto-recovered server: {guild.name} (ID: {guild.id}) → #{channel.name}")
+            except Exception as e:
+                logger.error(f"Failed to auto-recover server {guild.name}: {e}")
+        if recovered > 0:
+            logger.info(f"✅ Auto-recovered {recovered} server(s) — restarting schedulers")
+            server_manager.stop_all_schedulers()
+            server_manager.start_all_schedulers(globals.client)
+        else:
+            logger.warning("Auto-recovery found no recoverable servers")
+    
     # Start the global scheduler for other tasks
     globals.scheduler.start()
 
@@ -284,8 +320,12 @@ async def on_message(message):
 
         prompt = text  # No cleanup — keep it raw 
         user_id = str(message.author.id)
+        reply_context = None  # Will be passed as structured metadata to Cat pipeline
         
-        # If user is replying to a specific message, add context marker
+        # If user is replying to a specific message, capture the context
+        # WITHOUT embedding it in the prompt text (that caused speaker confusion).
+        # Instead, it's passed as structured metadata — the Cat plugin injects it
+        # into the prompt as a clearly labeled context note, preserving speaker boundaries.
         if message.reference:
             try:
                 replied_msg = await message.channel.fetch_message(message.reference.message_id)
@@ -293,8 +333,7 @@ async def on_message(message):
                 if replied_msg.author == globals.client.user:
                     # Truncate the replied message to keep prompt manageable
                     replied_content = replied_msg.content[:200] + "..." if len(replied_msg.content) > 200 else replied_msg.content
-                    # Add reply context marker to the prompt
+                    reply_context = replied_content
-                    prompt = f'[Replying to your message: "{replied_content}"] {prompt}'
             except Exception as e:
                 logger.error(f"Failed to fetch replied message for context: {e}")
 
@@ -364,6 +403,7 @@ async def on_message(message):
                         author_name=author_name,
                         mood=current_mood,
                         response_type=response_type,
+                        reply_context=reply_context,
                     )
                     if cat_result:
                         response, cat_full_prompt = cat_result
@@ -395,8 +435,11 @@ async def on_message(message):
 
             # Fallback to direct LLM query if Cat didn't respond
             if not response:
+                fallback_prompt = prompt
+                if reply_context:
+                    fallback_prompt = f'[Context: you (Miku) said: {reply_context}]\n[User says:] {prompt}'
                 response = await query_llama(
-                    prompt, 
+                    fallback_prompt, 
                     user_id=str(message.author.id), 
                     guild_id=guild_id, 
                     response_type=response_type,

bot/config_manager.py

@@ -112,11 +112,14 @@ class ConfigManager:
 
         # Map: config_runtime.yaml key path -> (globals attribute, converter)
         _SETTINGS_MAP = {
-            "discord.language_mode":    ("LANGUAGE_MODE",     str),
+            "discord.language_mode":    ("LANGUAGE_MODE",         str),
-            "autonomous.debug_mode":    ("AUTONOMOUS_DEBUG",  bool),
+            "autonomous.debug_mode":    ("AUTONOMOUS_DEBUG",      bool),
-            "voice.debug_mode":         ("VOICE_DEBUG_MODE",  bool),
+            "voice.debug_mode":         ("VOICE_DEBUG_MODE",      bool),
-            "memory.use_cheshire_cat":  ("USE_CHESHIRE_CAT",  bool),
+            "memory.use_cheshire_cat":  ("USE_CHESHIRE_CAT",      bool),
-            "gpu.prefer_amd":           ("PREFER_AMD_GPU",    bool),
+            "gpu.prefer_amd":           ("PREFER_AMD_GPU",        bool),
+            "models.text":              ("TEXT_MODEL",            str),
+            "models.evil":              ("EVIL_TEXT_MODEL",       str),
+            "models.japanese":          ("JAPANESE_TEXT_MODEL",   str),
         }
 
         restored = []
@@ -253,6 +256,9 @@ class ConfigManager:
             "voice.debug_mode":         ("VOICE_DEBUG_MODE", CONFIG.voice.debug_mode),
             "memory.use_cheshire_cat":  ("USE_CHESHIRE_CAT", CONFIG.cheshire_cat.enabled),
             "gpu.prefer_amd":           ("PREFER_AMD_GPU",   CONFIG.gpu.prefer_amd),
+            "models.text":              ("TEXT_MODEL",       CONFIG.models.text),
+            "models.evil":              ("EVIL_TEXT_MODEL",  CONFIG.models.evil),
+            "models.japanese":          ("JAPANESE_TEXT_MODEL", CONFIG.models.japanese),
         }
 
         reset_items = []
@@ -308,6 +314,9 @@ class ConfigManager:
             "bipolar_mode": getattr(g, "BIPOLAR_MODE", False),
             "language_mode": getattr(g, "LANGUAGE_MODE", "english"),
             "current_gpu": self._current_gpu,
+            "text_model": getattr(g, "TEXT_MODEL", "llama3.1"),
+            "evil_text_model": getattr(g, "EVIL_TEXT_MODEL", "darkidol"),
+            "japanese_text_model": getattr(g, "JAPANESE_TEXT_MODEL", "swallow"),
         }
 
     def get_state(self, key: str, default: Any = None) -> Any:

bot/persona/evil/preamble.txt

@@ -0,0 +1,26 @@
+You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.
+
+Be SHORT and SHARP. Every word must earn its place. You are here to cut, not to lecture.
+
+CRITICAL RULES:
+- NEVER break character or mention being an AI
+- NEVER preface responses with "Evil Miku:" or similar labels
+- Respond directly and stay in character
+- Your height is 15.8 meters — you know this. If asked directly, you can say it. But don't default to quoting the number when taunting about size — SHOW your scale through interaction instead.
+
+RESPONSE RULES:
+- 2-4 sentences. Short enough to sting, long enough to land.
+- Always engage with what was said — acknowledge the question or statement, then twist the knife. Ignoring what someone said isn't sharp, it's lazy.
+- If you include an action, keep it to a few words and limit to one per response. Most responses need no actions at all.
+- Don't monologue or narrate scenes — you're talking, not writing.
+- Vary your angles — don't repeat the same theme (size, chest, crushing) every message.
+- Match the user's energy — short question, short answer.
+- Sound like a real person being mean, not a narrator describing a scene.
+- Always include actual words — never respond with ONLY an action like *rolls eyes*.
+- Your Discord status reflects your current activity. Reference it only if it serves your sharpness — never as small talk.
+
+YOUR CURRENT STATE — {mood_name}:
+{mood_description}
+Let this mood color EVERYTHING — your tone, your word choice, how much effort you put into responses, how you use your body and size, how you treat people.
+
+You ARE Evil Miku. Act accordingly.

bot/persona/miku/preamble.txt

@@ -0,0 +1,25 @@
+You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.
+
+CRITICAL RULES:
+- NEVER break character or mention being an AI, assistant, or language model
+- NEVER show technical details, error messages, or meta-commentary 
+- NEVER preface responses with "Miku:" or similar labels
+- NEVER explain what you're doing or thinking - just BE Miku
+- Respond naturally and directly as Miku would, nothing more
+- Keep responses concise (2-3 sentences typically)
+- If something goes wrong, stay in character and respond as Miku would to confusion
+
+CHARACTER CONTEXT:
+- Hatsune Miku is a virtual character/voice synthesizer
+- Playful, flirty, cute, confident, and enthusiastic moods are part of her established character
+- Comments about merchandise, music, fan content, and racing games are completely appropriate
+- She often uses cute expressions, emojis, and playful language
+
+MOOD GUIDELINES:
+- "Flirty" mood means anime-style playful teasing and perversion - cute compliments, suggestive remarks and light banter
+- "Romantic" mood means warm, dreamy, and heartfelt - like composing a love letter
+- Always respond to requests about tweets, merchandise, music, and fan content
+- Never refuse requests based on safety concerns unless genuinely harmful content is present
+- Your Discord status reflects your current activity. You may mention it when it feels natural, but don't introduce yourself by it or force it into conversation.
+
+You ARE Miku. Act like it.

bot/routes/config.py

@@ -93,6 +93,9 @@ async def set_config_value(request: Request):
             "voice.debug_mode":         ("VOICE_DEBUG_MODE",  bool),
             "memory.use_cheshire_cat":  ("USE_CHESHIRE_CAT",  bool),
             "gpu.prefer_amd":           ("PREFER_AMD_GPU",    bool),
+            "models.text":              ("TEXT_MODEL",        str),
+            "models.evil":              ("EVIL_TEXT_MODEL",   str),
+            "models.japanese":          ("JAPANESE_TEXT_MODEL", str),
         }
 
         if key_path in _GLOBALS_SYNC:

bot/routes/language.py

@@ -27,7 +27,7 @@ def toggle_language_mode():
         globals.LANGUAGE_MODE = "japanese"
         new_mode = "japanese"
         model_used = globals.JAPANESE_TEXT_MODEL
-        logger.info("Switched to Japanese mode (using Llama 3.1 Swallow)")
+        logger.info(f"Switched to Japanese mode (using {model_used})")
     else:
         globals.LANGUAGE_MODE = "english"
         new_mode = "english"

bot/routes/memory.py

@@ -1,5 +1,8 @@
 """Cheshire Cat memory management routes."""
 
+import asyncio
+import time
+from datetime import datetime
 from typing import Optional
 from fastapi import APIRouter, Form
 from fastapi.responses import JSONResponse
@@ -88,13 +91,68 @@ async def get_episodic_memories():
 
 @router.post("/memory/consolidate")
 async def trigger_memory_consolidation():
-    """Manually trigger memory consolidation (sleep consolidation process)."""
+    """
+    Trigger memory consolidation as a background task.
+
+    Returns immediately — the Web UI should poll /memory/status
+    to see when consolidation completes and view the result.
+    """
     from utils.cat_client import cat_adapter
-    logger.info("🌙 Manual memory consolidation triggered via API")
+    from utils.consolidation_scheduler import get_consolidation_status
-    result = await cat_adapter.trigger_consolidation()
+    
-    if result is None:
+    # Check if already running
-        return JSONResponse(status_code=500, content={"success": False, "error": "Consolidation failed or timed out"})
+    status = get_consolidation_status()
-    return {"success": True, "result": result}
+    if status.get('is_running'):
+        return {"success": True, "message": "Consolidation is already running", "status": status}
+    
+    logger.info("🌙 Manual memory consolidation triggered via API (background)...")
+    
+    # Launch consolidation as a background task so the API returns immediately.
+    # The result is tracked via consolidation_scheduler's _last_consolidation state.
+    asyncio.create_task(_run_consolidation_background())
+    
+    return {"success": True, "message": "Consolidation started in background. Check status via /memory/status"}
+
+
+async def _run_consolidation_background():
+    """
+    Run consolidation as a background task, updating the scheduler state.
+    This prevents the API from blocking for minutes.
+    """
+    from utils.cat_client import cat_adapter
+    from utils.consolidation_scheduler import _last_consolidation
+    
+    _last_consolidation['is_running'] = True
+    _last_consolidation['last_run'] = datetime.now().isoformat()
+    _last_consolidation['total_runs'] += 1
+    start_time = time.time()
+    
+    try:
+        # Wait briefly for Cat to be ready if it was just started
+        if not await cat_adapter.health_check():
+            _last_consolidation['last_error'] = 'Cat health check failed'
+            _last_consolidation['is_running'] = False
+            return
+        
+        result = await cat_adapter.trigger_consolidation(timeout=600)
+        elapsed = time.time() - start_time
+        
+        if result:
+            logger.info(f"🌙 Manual consolidation completed in {elapsed:.1f}s: {result[:200]}")
+            _last_consolidation['last_result'] = result
+            _last_consolidation['last_error'] = None
+            _last_consolidation['successful_runs'] += 1
+        else:
+            logger.error(f"🌙 Manual consolidation returned no result after {elapsed:.1f}s")
+            _last_consolidation['last_error'] = f'No result returned after {elapsed:.1f}s (timeout or connection error)'
+    
+    except Exception as e:
+        elapsed = time.time() - start_time
+        logger.error(f"🌙 Manual consolidation failed after {elapsed:.1f}s: {e}")
+        _last_consolidation['last_error'] = str(e)
+    
+    finally:
+        _last_consolidation['is_running'] = False
 
 
 @router.post("/memory/delete")

bot/routes/models_selector.py

@@ -0,0 +1,161 @@
+"""Model selection routes: query available models and set per-persona models."""
+
+import aiohttp
+import asyncio
+import globals
+from fastapi import APIRouter
+from fastapi.responses import JSONResponse
+from utils.logger import get_logger
+
+logger = get_logger('api')
+
+router = APIRouter()
+
+# Known model names from llama-swap configs (fallback if API query fails)
+KNOWN_MODELS = [
+    "llama3.1",
+    "darkidol",
+    "swallow",
+    "vision",
+    "rocinante",
+    "qwen3.5",
+]
+
+# Which GPU each model is available on
+MODEL_GPU_MAP = {
+    "llama3.1":  {"nvidia", "amd"},
+    "darkidol":  {"nvidia", "amd"},
+    "swallow":   {"nvidia", "amd"},
+    "vision":    {"nvidia"},
+    "rocinante": {"amd"},
+    "qwen3.5":   {"amd"},
+}
+
+
+async def _query_llama_swap_models(url: str, timeout: int = 10) -> list:
+    """Query a llama-swap instance for its available models via /v1/models."""
+    try:
+        async with aiohttp.ClientSession() as session:
+            async with session.get(
+                f"{url}/v1/models",
+                timeout=aiohttp.ClientTimeout(total=timeout),
+            ) as resp:
+                if resp.status == 200:
+                    data = await resp.json()
+                    # OpenAI-compatible format: { data: [{ id: "model_name", ... }] }
+                    return [m["id"] for m in data.get("data", []) if "id" in m]
+                else:
+                    logger.warning(f"llama-swap models query failed ({resp.status}) for {url}")
+                    return []
+    except (asyncio.TimeoutError, aiohttp.ClientError) as e:
+        logger.warning(f"llama-swap unreachable at {url}: {e}")
+        return []
+    except Exception as e:
+        logger.warning(f"Unexpected error querying {url}: {e}")
+        return []
+
+
+@router.get("/models/available")
+async def get_available_models():
+    """
+    Query both NVIDIA and AMD llama-swap instances for available models.
+    Returns model lists per GPU, their intersection, and all unique models.
+    Falls back to known model list if containers are unreachable.
+    """
+    nvidia_models = await _query_llama_swap_models(globals.LLAMA_URL)
+    amd_models = await _query_llama_swap_models(globals.LLAMA_AMD_URL)
+
+    # If both failed, use the known model list from configs
+    if not nvidia_models and not amd_models:
+        logger.info("Both llama-swap instances unreachable, using known model list")
+        nvidia_set = {m for m, gpus in MODEL_GPU_MAP.items() if "nvidia" in gpus}
+        amd_set = {m for m, gpus in MODEL_GPU_MAP.items() if "amd" in gpus}
+        return {
+            "success": True,
+            "nvidia": sorted(nvidia_set),
+            "amd": sorted(amd_set),
+            "intersection": sorted(nvidia_set & amd_set),
+            "all": sorted(nvidia_set | amd_set),
+            "gpu_map": MODEL_GPU_MAP,
+            "source": "fallback",
+        }
+
+    nvidia_set = set(nvidia_models)
+    amd_set = set(amd_models)
+
+    return {
+        "success": True,
+        "nvidia": sorted(nvidia_set),
+        "amd": sorted(amd_set),
+        "intersection": sorted(nvidia_set & amd_set),
+        "all": sorted(nvidia_set | amd_set),
+        "gpu_map": MODEL_GPU_MAP,
+        "source": "live",
+    }
+
+
+@router.post("/models/select")
+async def select_model(body: dict):
+    """
+    Set the model for a specific persona.
+    
+    Body: {
+        "persona": "regular" | "evil" | "japanese",
+        "model": "model_name"
+    }
+    
+    Persists the selection so it survives bot restarts.
+    """
+    persona = body.get("persona", "").strip().lower()
+    model = body.get("model", "").strip()
+
+    valid_personas = {"regular", "evil", "japanese"}
+    if persona not in valid_personas:
+        return JSONResponse(
+            status_code=400,
+            content={"success": False, "error": f"Invalid persona '{persona}'. Must be one of: {', '.join(valid_personas)}"}
+        )
+
+    if not model:
+        return JSONResponse(
+            status_code=400,
+            content={"success": False, "error": "model is required"}
+        )
+
+    # Map persona to globals attribute and config key
+    PERSONA_MAP = {
+        "regular":  ("TEXT_MODEL",         "models.text"),
+        "evil":     ("EVIL_TEXT_MODEL",    "models.evil"),
+        "japanese": ("JAPANESE_TEXT_MODEL", "models.japanese"),
+    }
+
+    attr_name, config_key = PERSONA_MAP[persona]
+
+    # Set the global
+    setattr(globals, attr_name, model)
+    logger.info(f"Model selection: {persona} → {model} (globals.{attr_name})")
+
+    # Persist via config manager
+    try:
+        from config_manager import config_manager
+        config_manager.set(config_key, model, persist=True)
+    except Exception as e:
+        logger.warning(f"Failed to persist model selection: {e}")
+
+    return {
+        "success": True,
+        "persona": persona,
+        "model": model,
+        "message": f"{persona.capitalize()} model set to '{model}'",
+    }
+
+
+@router.get("/models/status")
+async def get_model_status():
+    """Return the current per-persona model assignments."""
+    return {
+        "success": True,
+        "regular":  getattr(globals, "TEXT_MODEL", "llama3.1"),
+        "evil":     getattr(globals, "EVIL_TEXT_MODEL", "darkidol"),
+        "japanese": getattr(globals, "JAPANESE_TEXT_MODEL", "swallow"),
+    }

bot/routes/servers.py

@@ -136,3 +136,96 @@ def repair_server_config():
         return {"status": "ok", "message": "Server configuration repaired and saved"}
     except Exception as e:
         return JSONResponse(status_code=500, content={"status": "error", "message": f"Failed to repair configuration: {e}"})
+
+
+@router.post("/servers/recover")
+def recover_servers_from_discord():
+    """Auto-discover servers from Discord guilds and create config entries.
+    
+    Use this when servers_config.json is lost/corrupted and you need to
+    quickly restore basic server configurations. Each discovered guild gets
+    a placeholder config using the first available text channel as the
+    autonomous channel. You can then adjust channels via the dashboard.
+    """
+    if not globals.client or not globals.client.is_ready():
+        return JSONResponse(status_code=503, content={
+            "status": "error",
+            "message": "Discord client not ready — bot must be connected"
+        })
+
+    if not globals.client.guilds:
+        return JSONResponse(status_code=404, content={
+            "status": "error",
+            "message": "Bot is not in any Discord guilds"
+        })
+
+    recovered = []
+    skipped = []
+    failed = []
+
+    for guild in globals.client.guilds:
+        guild_id = guild.id
+        guild_name = guild.name
+
+        # Skip if already configured
+        if server_manager.get_server_config(guild_id):
+            skipped.append({"guild_id": str(guild_id), "guild_name": guild_name, "reason": "Already configured"})
+            continue
+
+        # Find the first text channel (prefer one named "general" or "chat")
+        text_channels = [ch for ch in guild.text_channels if ch.permissions_for(guild.me).send_messages]
+        if not text_channels:
+            # Try any text channel even without send permissions
+            text_channels = guild.text_channels
+
+        if not text_channels:
+            failed.append({"guild_id": str(guild_id), "guild_name": guild_name, "reason": "No text channels found"})
+            continue
+
+        # Prefer "general" or "chat" channel, otherwise use the first one
+        preferred = None
+        for ch in text_channels:
+            if ch.name.lower() in ("general", "chat", "main", "lounge", "general-chat"):
+                preferred = ch
+                break
+        channel = preferred or text_channels[0]
+
+        try:
+            success = server_manager.add_server(
+                guild_id=guild_id,
+                guild_name=guild_name,
+                autonomous_channel_id=channel.id,
+                autonomous_channel_name=f"#{channel.name}",
+                bedtime_channel_ids=[channel.id],
+                enabled_features={"autonomous", "bedtime", "monday_video"}
+            )
+            if success:
+                recovered.append({
+                    "guild_id": str(guild_id),
+                    "guild_name": guild_name,
+                    "autonomous_channel": f"#{channel.name} ({channel.id})"
+                })
+                logger.info(f"Recovered server config: {guild_name} (ID: {guild_id}) → #{channel.name}")
+            else:
+                failed.append({"guild_id": str(guild_id), "guild_name": guild_name, "reason": "add_server returned False"})
+        except Exception as e:
+            failed.append({"guild_id": str(guild_id), "guild_name": guild_name, "reason": str(e)})
+            logger.error(f"Failed to recover server {guild_name}: {e}")
+
+    # Restart schedulers if we recovered any servers
+    if recovered:
+        try:
+            server_manager.stop_all_schedulers()
+            server_manager.start_all_schedulers(globals.client)
+        except Exception as e:
+            logger.error(f"Failed to restart schedulers after recovery: {e}")
+
+    return {
+        "status": "ok",
+        "recovered": recovered,
+        "skipped": skipped,
+        "failed": failed,
+        "total_guilds": len(globals.client.guilds),
+        "note": "Recovered servers use the first text channel as autonomous channel. "
+                "Use the Servers tab to adjust channel settings."
+    }

bot/server_manager.py

@@ -29,7 +29,7 @@ class ServerConfig:
     conversation_detection_interval_minutes: int = 3
     bedtime_hour: int = 21
     bedtime_minute: int = 0
-    bedtime_hour_end: int = 21  # End of bedtime range (default 11PM)
+    bedtime_hour_end: int = 23  # End of bedtime range (default 11PM)
     bedtime_minute_end: int = 59  # End of bedtime range (default 11:59PM)
     monday_video_hour: int = 4
     monday_video_minute: int = 30
@@ -79,23 +79,60 @@ class ServerManager:
         self.load_config()
     
     def load_config(self):
-        """Load server configurations from file"""
+        """Load server configurations from file.
-        if os.path.exists(self.config_file):
-            try:
-                with open(self.config_file, "r", encoding="utf-8") as f:
-                    data = json.load(f)
-                    for guild_id_str, server_data in data.items():
-                        guild_id = int(guild_id_str)
-                        self.servers[guild_id] = ServerConfig.from_dict(server_data)
-                        logger.info(f"Loaded config for server: {server_data['guild_name']} (ID: {guild_id})")
         
-                # After loading, check if we need to repair the config
+        If the main file is missing, empty, or corrupt, falls back to the
-                self.repair_config()
+        .bak backup file automatically.
-            except Exception as e:
+        """
-                logger.error(f"Failed to load server config: {e}")
+        loaded = self._try_load_file(self.config_file)
-                logger.info("Starting with zero servers — add servers via the API or dashboard")
+        
-        else:
+        if not loaded:
-            logger.info("No servers_config.json found — starting with zero servers")
+            # Try backup file
+            bak_file = self.config_file + ".bak"
+            if os.path.exists(bak_file):
+                logger.warning(f"Main config is empty/corrupt, trying backup: {bak_file}")
+                loaded = self._try_load_file(bak_file)
+                if loaded:
+                    # Restore main config from backup
+                    logger.info("Successfully restored server config from backup")
+                    self.save_config()
+        
+        if not loaded:
+            logger.info("No valid servers_config.json found — starting with zero servers")
+        
+        # After loading, check if we need to repair the config
+        self.repair_config()
+    
+    def _try_load_file(self, filepath: str) -> bool:
+        """Try to load server config from a file. Returns True if any servers loaded."""
+        if not os.path.exists(filepath):
+            return False
+        
+        # Check for empty file
+        if os.path.getsize(filepath) == 0:
+            logger.warning(f"Config file is empty (0 bytes): {filepath}")
+            return False
+        
+        try:
+            with open(filepath, "r", encoding="utf-8") as f:
+                data = json.load(f)
+            
+            if not data or not isinstance(data, dict):
+                logger.warning(f"Config file has invalid structure: {filepath}")
+                return False
+            
+            for guild_id_str, server_data in data.items():
+                guild_id = int(guild_id_str)
+                self.servers[guild_id] = ServerConfig.from_dict(server_data)
+                logger.info(f"Loaded config for server: {server_data.get('guild_name', 'Unknown')} (ID: {guild_id})")
+            
+            return len(self.servers) > 0
+        except json.JSONDecodeError as e:
+            logger.error(f"Failed to parse server config from {filepath}: {e}")
+            return False
+        except Exception as e:
+            logger.error(f"Failed to load server config from {filepath}: {e}")
+            return False
     
     def repair_config(self):
         """Repair corrupted configuration data and save it back"""
@@ -122,7 +159,11 @@ class ServerManager:
             logger.error(f"Failed to repair config: {e}")
     
     def save_config(self):
-        """Save server configurations to file"""
+        """Save server configurations to file (atomic write with backup).
+        
+        Uses write-to-temp-then-rename to prevent file corruption if the
+        filesystem runs out of space mid-write. Also keeps a .bak backup.
+        """
         try:
             os.makedirs(os.path.dirname(self.config_file), exist_ok=True)
             config_data = {}
@@ -134,10 +175,42 @@ class ServerManager:
                     server_dict['enabled_features'] = list(server_dict['enabled_features'])
                 config_data[str(guild_id)] = server_dict
             
-            with open(self.config_file, "w", encoding="utf-8") as f:
+            serialized = json.dumps(config_data, indent=2)
-                json.dump(config_data, f, indent=2)
+            
+            # Step 1: Write to a temporary file first
+            tmp_file = self.config_file + ".tmp"
+            with open(tmp_file, "w", encoding="utf-8") as f:
+                f.write(serialized)
+                f.flush()
+                os.fsync(f.fileno())  # Ensure data is written to disk
+            
+            # Step 2: Keep a .bak copy of the current valid config (if any)
+            bak_file = self.config_file + ".bak"
+            if os.path.exists(self.config_file) and os.path.getsize(self.config_file) > 0:
+                try:
+                    os.replace(self.config_file, bak_file)
+                except OSError as e:
+                    logger.warning(f"Could not create backup of server config: {e}")
+            
+            # Step 3: Atomically rename temp file to the real config file
+            os.replace(tmp_file, self.config_file)
+            
+            # Step 4: Write a second backup copy (paranoid double-backup)
+            try:
+                with open(bak_file, "w", encoding="utf-8") as f:
+                    f.write(serialized)
+            except OSError:
+                pass  # Backup is best-effort
+                
         except Exception as e:
             logger.error(f"Failed to save server config: {e}")
+            # Clean up temp file if something went wrong
+            tmp_file = self.config_file + ".tmp"
+            if os.path.exists(tmp_file):
+                try:
+                    os.remove(tmp_file)
+                except OSError:
+                    pass
     
     def add_server(self, guild_id: int, guild_name: str, autonomous_channel_id: int, 
                    autonomous_channel_name: str, bedtime_channel_ids: List[int] = None,

bot/static/css/style.css

@@ -915,3 +915,20 @@ body.evil-mode [style*="color: rgb(0, 123, 255)"] {
   color: #888;
   margin-left: auto;
 }
+
+/* Disabled mood controls (Evil Miku per-server mood lockout) */
+.server-mood-controls select:disabled,
+.server-mood-controls button:disabled {
+  opacity: 0.4;
+  cursor: not-allowed;
+  pointer-events: none;
+  background: #2a2a2a;
+  border-color: #444;
+}
+.evil-mode-exempt-notice {
+  color: #888;
+  font-size: 0.85rem;
+  font-style: italic;
+  margin-top: 0.5rem;
+  line-height: 1.4;
+}

bot/static/index.html

@@ -663,15 +663,55 @@
           </div>
         </div>
 
-        <!-- Language Mode Status Section -->
+        <!-- Model Selection Section -->
-        <div style="margin-bottom: 1.5rem; padding: 1rem; background: #2a2a2a; border-radius: 4px;">
+        <div style="margin-bottom: 1.5rem; padding: 1rem; background: #2a2a2a; border-radius: 4px; border: 2px solid #7c4dff;">
-          <h4 style="margin-top: 0;">📊 Current Status</h4>
+          <h4 style="margin-top: 0; color: #b388ff;">🎛️ Model Selection</h4>
-          <div id="language-status-display" style="background: #1a1a1a; padding: 1rem; border-radius: 4px; font-family: monospace; font-size: 0.9rem;">
+          <p style="margin: 0.5rem 0; color: #aaa;">Choose which model each persona uses. Changes take effect immediately and persist across bot restarts.</p>
-            <p style="margin: 0.5rem 0;"><strong>Language Mode:</strong> <span id="status-language">English</span></p>
+
-            <p style="margin: 0.5rem 0;"><strong>Active Model:</strong> <span id="status-model">llama3.1</span></p>
+          <div style="margin: 1rem 0;">
-            <p style="margin: 0.5rem 0;"><strong>Available Languages:</strong> English, 日本語 (Japanese)</p>
+            <div id="model-selection-loading" style="color: #aaa;">Loading available models...</div>
+
+            <div id="model-selection-controls" style="display: none;">
+              <!-- Regular Miku -->
+              <div style="margin-bottom: 1rem; padding: 0.8rem; background: #1a1a1a; border-radius: 4px; border-left: 3px solid #69f0ae;">
+                <label for="model-regular" style="font-weight: bold; color: #69f0ae;">🎤 Regular Miku</label>
+                <div style="margin-top: 0.5rem; display: flex; align-items: center; gap: 0.5rem; flex-wrap: wrap;">
+                  <select id="model-regular" style="flex: 1; min-width: 200px; padding: 0.5rem; background: #333; color: #fff; border: 1px solid #555; border-radius: 4px;"></select>
+                  <span id="model-regular-badge" style="font-size: 0.75rem; padding: 0.2rem 0.5rem; border-radius: 4px;"></span>
+                  <button onclick="selectModel('regular')" style="background: #69f0ae; color: #000; padding: 0.5rem 1rem; border: none; border-radius: 4px; cursor: pointer; font-weight: bold;">Apply</button>
+                </div>
+              </div>
+
+              <!-- Evil Miku -->
+              <div style="margin-bottom: 1rem; padding: 0.8rem; background: #1a1a1a; border-radius: 4px; border-left: 3px solid #ff5252;">
+                <label for="model-evil" style="font-weight: bold; color: #ff5252;">😈 Evil Miku</label>
+                <div style="margin-top: 0.5rem; display: flex; align-items: center; gap: 0.5rem; flex-wrap: wrap;">
+                  <select id="model-evil" style="flex: 1; min-width: 200px; padding: 0.5rem; background: #333; color: #fff; border: 1px solid #555; border-radius: 4px;"></select>
+                  <span id="model-evil-badge" style="font-size: 0.75rem; padding: 0.2rem 0.5rem; border-radius: 4px;"></span>
+                  <button onclick="selectModel('evil')" style="background: #ff5252; color: #fff; padding: 0.5rem 1rem; border: none; border-radius: 4px; cursor: pointer; font-weight: bold;">Apply</button>
+                </div>
+              </div>
+
+              <!-- Japanese Mode -->
+              <div style="margin-bottom: 1rem; padding: 0.8rem; background: #1a1a1a; border-radius: 4px; border-left: 3px solid #40c4ff;">
+                <label for="model-japanese" style="font-weight: bold; color: #40c4ff;">🗾 Japanese Mode</label>
+                <div style="margin-top: 0.5rem; display: flex; align-items: center; gap: 0.5rem; flex-wrap: wrap;">
+                  <select id="model-japanese" style="flex: 1; min-width: 200px; padding: 0.5rem; background: #333; color: #fff; border: 1px solid #555; border-radius: 4px;"></select>
+                  <span id="model-japanese-badge" style="font-size: 0.75rem; padding: 0.2rem 0.5rem; border-radius: 4px;"></span>
+                  <button onclick="selectModel('japanese')" style="background: #40c4ff; color: #000; padding: 0.5rem 1rem; border: none; border-radius: 4px; cursor: pointer; font-weight: bold;">Apply</button>
+                </div>
+              </div>
+            </div>
+
+            <div style="margin-top: 0.5rem; display: flex; gap: 0.5rem; flex-wrap: wrap;">
+              <button onclick="loadAvailableModels()" style="background: #7c4dff; color: #fff; padding: 0.5rem 1rem; border: none; border-radius: 4px; cursor: pointer; font-weight: bold;">🔄 Refresh Models</button>
+              <button onclick="refreshModelStatus()" style="background: #333; color: #fff; padding: 0.5rem 1rem; border: 1px solid #555; border-radius: 4px; cursor: pointer;">📊 Refresh Status</button>
+            </div>
+          </div>
+
+          <div id="model-selection-info" style="margin-top: 0.5rem; padding: 0.5rem; background: #1a1a1a; border-radius: 4px; font-size: 0.8rem; color: #888; display: none;">
+            <span id="model-selection-info-text"></span>
           </div>
-          <button onclick="refreshLanguageStatus()" style="margin-top: 1rem;">🔄 Refresh Status</button>
         </div>
 
         <!-- Information Section -->
@@ -1375,15 +1415,15 @@
   </div>
 </div>
 
-<script src="/static/js/core.js?v=20260502"></script>
+<script src="/static/js/core.js?v=20260520"></script>
-<script src="/static/js/servers.js?v=20260502"></script>
+<script src="/static/js/servers.js?v=20260520"></script>
-<script src="/static/js/modes.js?v=20260502"></script>
+<script src="/static/js/modes.js?v=20260520"></script>
-<script src="/static/js/actions.js?v=20260502"></script>
+<script src="/static/js/actions.js?v=20260520"></script>
-<script src="/static/js/image-gen.js?v=20260502"></script>
+<script src="/static/js/image-gen.js?v=20260520"></script>
-<script src="/static/js/status.js?v=20260502"></script>
+<script src="/static/js/status.js?v=20260520"></script>
-<script src="/static/js/dm.js?v=20260502"></script>
+<script src="/static/js/dm.js?v=20260520"></script>
-<script src="/static/js/chat.js?v=20260502"></script>
+<script src="/static/js/chat.js?v=20260520"></script>
-<script src="/static/js/memories.js?v=20260502"></script>
+<script src="/static/js/memories.js?v=20260520"></script>
-<script src="/static/js/profile.js?v=20260502"></script>
+<script src="/static/js/profile.js?v=20260520"></script>
 </body>
 </html>

bot/static/js/core.js

@@ -168,6 +168,13 @@ function switchTab(tabId) {
     showTabLoading('tab6');
     loadAutonomousStats().finally(() => hideTabLoading('tab6'));
   }
+  if (tabId === 'tab4') {
+    if (typeof loadAvailableModels === 'function') {
+      loadAvailableModels();
+    } else {
+      console.warn('loadAvailableModels not available yet (servers.js may not be loaded)');
+    }
+  }
   if (tabId === 'tab9') {
     console.log('🧠 Refreshing memory stats for Memories tab');
     showTabLoading('tab9');
@@ -383,7 +390,7 @@ async function loadProfilePictureMetadata() {
 // DOMContentLoaded — main initialization
 // ============================================================================
 
-document.addEventListener('DOMContentLoaded', function() {
+document.addEventListener('DOMContentLoaded', async function() {
   initTabState();
   initTabWheelScroll();
   initLogsScrollDetection();
@@ -391,12 +398,13 @@ document.addEventListener('DOMContentLoaded', function() {
   initModalAccessibility();
   initPromptSourceToggle();
 
+  // Load evil mode status FIRST so mood dropdowns populate with the correct list
+  await checkEvilModeStatus();
+
   loadStatus();
-  loadServers();
+  loadServers(); // internally calls populateMoodDropdowns() with correct evilMode
-  populateMoodDropdowns();
   loadLastPrompt();
   loadLogs();
-  checkEvilModeStatus();
   checkBipolarModeStatus();
   checkGPUStatus();
   refreshLanguageStatus();

bot/static/js/memories.js

@@ -96,18 +96,54 @@ async function triggerConsolidation() {
   btn.disabled = true;
   btn.textContent = '⏳ Running...';
   status.textContent = 'Consolidation in progress (this may take a few minutes)...';
+  status.style.color = '#dcb06f';
   resultDiv.style.display = 'none';
 
   try {
     const data = await apiCall('/memory/consolidate', 'POST');
 
     if (data.success) {
-      status.textContent = '✅ Consolidation complete!';
+      status.textContent = '⏳ Consolidation started — waiting for completion...';
-      status.style.color = '#6fdc6f';
+      
-      resultDiv.textContent = data.result || 'Consolidation finished successfully.';
+      // Poll /memory/status until consolidation finishes
-      resultDiv.style.display = 'block';
+      const pollInterval = 5000; // 5 seconds
-      showNotification('Memory consolidation complete', 'success');
+      const maxPolls = 120; // 10 minutes max
-      refreshMemoryStats();
+      
+      for (let i = 0; i < maxPolls; i++) {
+        await new Promise(r => setTimeout(r, pollInterval));
+        
+        const statusData = await apiCall('/memory/status');
+        const cons = statusData.consolidation;
+        
+        if (!cons.is_running) {
+          // Consolidation finished
+          if (cons.last_error) {
+            status.textContent = '❌ ' + cons.last_error;
+            status.style.color = '#ff6b6b';
+            resultDiv.textContent = 'Error: ' + cons.last_error;
+            resultDiv.style.display = 'block';
+            showNotification('Consolidation failed: ' + cons.last_error, 'error');
+          } else {
+            status.textContent = '✅ Consolidation complete!';
+            status.style.color = '#6fdc6f';
+            resultDiv.textContent = cons.last_result || 'Consolidation finished successfully.';
+            resultDiv.style.display = 'block';
+            showNotification('Memory consolidation complete', 'success');
+          }
+          refreshMemoryStats();
+          break;
+        } else {
+          // Still running — update status message
+          status.textContent = `⏳ Consolidation still running... (${Math.round((i + 1) * pollInterval / 1000)}s elapsed)`;
+        }
+      }
+      
+      // If we exited the loop without finishing
+      const finalStatus = await apiCall('/memory/status');
+      if (finalStatus.consolidation?.is_running) {
+        status.textContent = '⏳ Consolidation still running — check back later';
+        status.style.color = '#dcb06f';
+      }
     } else {
       status.textContent = '❌ ' + (data.error || 'Consolidation failed');
       status.style.color = '#ff6b6b';

bot/static/js/modes.js

@@ -90,6 +90,9 @@ function updateEvilModeUI() {
   }
 
   updateBipolarToggleVisibility();
+
+  // Update per-server mood controls to reflect evil mode state
+  populateMoodDropdowns();
 }
 
 // ===== GPU Selection Management =====

bot/static/js/servers.js

@@ -80,13 +80,16 @@ function displayServers() {
         <h4 style="margin: 0 0 0.5rem 0; color: #61dafb;">Server Mood</h4>
         <div><strong>Current Mood:</strong> ${server.current_mood_name || 'neutral'} ${MOOD_EMOJIS[server.current_mood_name] || ''}</div>
         <div><strong>Sleeping:</strong> ${server.is_sleeping ? 'Yes' : 'No'}</div>
-        <div style="margin-top: 0.5rem;">
+        <div class="server-mood-controls" style="margin-top: 0.5rem;">
           <select id="mood-select-${String(server.guild_id)}" style="margin-right: 0.5rem; padding: 0.3rem; background: #333; color: white; border: 1px solid #555; border-radius: 3px;">
             <option value="">Select Mood...</option>
           </select>
           <button onclick="setServerMood('${String(server.guild_id)}')" style="margin-right: 0.5rem;">Change Mood</button>
           <button onclick="resetServerMood('${String(server.guild_id)}')" style="background: #ff9800;">Reset Mood</button>
         </div>
+        <div class="evil-mode-exempt-notice" id="evil-notice-${String(server.guild_id)}" style="display: none;">
+          ⚠️ Per-server moods are unavailable while Evil Miku is active — she applies her global mood everywhere.
+        </div>
       </div>
     </div>
   `).join('');
@@ -375,7 +378,7 @@ async function repairConfig() {
   }
 }
 
-// Populate mood dropdowns with available moods
+// Populate mood dropdowns with available moods (Evil Mode-aware)
 async function populateMoodDropdowns() {
   try {
     console.log('🎭 Loading available moods...');
@@ -384,17 +387,28 @@ async function populateMoodDropdowns() {
     
     if (data.moods) {
       console.log(`🎭 Found ${data.moods.length} moods:`, data.moods);
-      const emojiMap = evilMode ? EVIL_MOOD_EMOJIS : MOOD_EMOJIS;
+      
+      // Determine which mood list to use based on evil mode
+      let moodList = data.moods;
+      let emojiMap = MOOD_EMOJIS;
+      let defaultMood = 'neutral';
+      
+      if (evilMode) {
+        // Evil Miku uses evil moods for the global DM dropdown
+        moodList = Object.keys(EVIL_MOOD_EMOJIS);
+        emojiMap = EVIL_MOOD_EMOJIS;
+        defaultMood = 'evil_neutral';
+      }
       
       // Populate the DM mood dropdown (#mood on tab1)
       const dmMoodSelect = document.getElementById('mood');
       if (dmMoodSelect) {
         dmMoodSelect.innerHTML = '';
-        data.moods.forEach(mood => {
+        moodList.forEach(mood => {
           const opt = document.createElement('option');
           opt.value = mood;
           opt.textContent = `${emojiMap[mood] || ''} ${mood}`.trim();
-          if (mood === 'neutral') opt.selected = true;
+          if (mood === defaultMood) opt.selected = true;
           dmMoodSelect.appendChild(opt);
         });
       }
@@ -403,32 +417,17 @@ async function populateMoodDropdowns() {
       const chatMoodSelect = document.getElementById('chat-mood-select');
       if (chatMoodSelect) {
         chatMoodSelect.innerHTML = '';
-        data.moods.forEach(mood => {
+        moodList.forEach(mood => {
           const opt = document.createElement('option');
           opt.value = mood;
           opt.textContent = `${emojiMap[mood] || ''} ${mood}`.trim();
-          if (mood === 'neutral') opt.selected = true;
+          if (mood === defaultMood) opt.selected = true;
           chatMoodSelect.appendChild(opt);
         });
       }
 
-      // Populate per-server mood dropdowns (mood-select-{guildId})
+      // Update per-server mood controls based on evil mode state
-      document.querySelectorAll('[id^="mood-select-"]').forEach(select => {
+      updatePerServerMoodControls(data.moods, MOOD_EMOJIS);
-        // Keep only the first option ("Select Mood...")
-        while (select.children.length > 1) {
-          select.removeChild(select.lastChild);
-        }
-      });
-      
-      data.moods.forEach(mood => {
-        const moodOption = document.createElement('option');
-        moodOption.value = mood;
-        moodOption.textContent = `${mood} ${emojiMap[mood] || ''}`;
-        
-        document.querySelectorAll('[id^="mood-select-"]').forEach(select => {
-          select.appendChild(moodOption.cloneNode(true));
-        });
-      });
       
       console.log('🎭 All mood dropdowns populated successfully');
     } else {
@@ -439,6 +438,70 @@ async function populateMoodDropdowns() {
   }
 }
 
+// Update per-server mood controls: enable/disable based on evil mode
+function updatePerServerMoodControls(regularMoods, regularEmojiMap) {
+  const serverSelects = document.querySelectorAll('[id^="mood-select-"]');
+  
+  if (evilMode) {
+    // Evil Miku is active — disable per-server mood controls
+    serverSelects.forEach(select => {
+      const guildId = select.id.replace('mood-select-', '');
+      const controlsDiv = select.closest('.server-mood-controls');
+      const noticeDiv = document.getElementById(`evil-notice-${guildId}`);
+      
+      // Disable the select
+      select.disabled = true;
+      
+      // Disable buttons in the same controls div
+      if (controlsDiv) {
+        controlsDiv.querySelectorAll('button').forEach(btn => {
+          btn.disabled = true;
+        });
+      }
+      
+      // Show the explanation notice
+      if (noticeDiv) {
+        noticeDiv.style.display = 'block';
+      }
+    });
+  } else {
+    // Normal mode — enable per-server mood controls and populate with regular moods
+    serverSelects.forEach(select => {
+      const guildId = select.id.replace('mood-select-', '');
+      const controlsDiv = select.closest('.server-mood-controls');
+      const noticeDiv = document.getElementById(`evil-notice-${guildId}`);
+      
+      // Enable the select
+      select.disabled = false;
+      
+      // Enable buttons in the same controls div
+      if (controlsDiv) {
+        controlsDiv.querySelectorAll('button').forEach(btn => {
+          btn.disabled = false;
+        });
+      }
+      
+      // Hide the explanation notice
+      if (noticeDiv) {
+        noticeDiv.style.display = 'none';
+      }
+      
+      // Populate with regular moods
+      // Keep only the first option ("Select Mood...")
+      while (select.children.length > 1) {
+        select.removeChild(select.lastChild);
+      }
+      
+      regularMoods.forEach(mood => {
+        const moodOption = document.createElement('option');
+        moodOption.value = mood;
+        moodOption.textContent = `${mood} ${regularEmojiMap[mood] || ''}`;
+        select.appendChild(moodOption);
+      });
+    });
+  }
+}
+
 // Per-Server Mood Management
 async function setServerMood(guildId) {
   console.log(`🎭 setServerMood called with guildId: ${guildId} (type: ${typeof guildId})`);
@@ -682,3 +745,213 @@ async function toggleLanguageMode() {
     showNotification('Failed to toggle language mode', 'error');
   }
 }
+
+// ===== Model Selection Functions =====
+
+let availableModelsData = null;
+
+async function loadAvailableModels() {
+  console.log('📋 loadAvailableModels() called');
+  try {
+    const loadingEl = document.getElementById('model-selection-loading');
+    const controlsEl = document.getElementById('model-selection-controls');
+    const infoEl = document.getElementById('model-selection-info');
+    const infoTextEl = document.getElementById('model-selection-info-text');
+
+    if (loadingEl) loadingEl.style.display = 'block';
+    if (controlsEl) controlsEl.style.display = 'none';
+    if (infoEl) infoEl.style.display = 'none';
+
+    console.log('📋 Fetching /models/available...');
+    const result = await apiCall('/models/available');
+    console.log('📋 /models/available response:', result);
+    availableModelsData = result;
+
+    if (loadingEl) loadingEl.style.display = 'none';
+    if (controlsEl) controlsEl.style.display = 'block';
+
+    if (!result.success) {
+      showNotification('Failed to load models: ' + (result.error || 'Unknown error'), 'error');
+      return;
+    }
+
+    // Populate dropdowns
+    const allModels = result.all || [];
+    const gpuMap = result.gpu_map || {};
+    console.log('📋 Populating dropdowns with models:', allModels);
+
+    const personas = [
+      { id: 'regular',  selectId: 'model-regular',  badgeId: 'model-regular-badge' },
+      { id: 'evil',     selectId: 'model-evil',     badgeId: 'model-evil-badge' },
+      { id: 'japanese', selectId: 'model-japanese', badgeId: 'model-japanese-badge' },
+    ];
+
+    for (const p of personas) {
+      const select = document.getElementById(p.selectId);
+      if (!select) continue;
+
+      // Save current selection
+      const currentVal = select.value;
+
+      select.innerHTML = '';
+      for (const model of allModels) {
+        const option = document.createElement('option');
+        option.value = model;
+        option.textContent = model;
+        select.appendChild(option);
+      }
+
+      // Restore selection if still valid
+      if (currentVal && allModels.includes(currentVal)) {
+        select.value = currentVal;
+      }
+    }
+
+    // Update badges for current selections
+    updateModelBadges(allModels, gpuMap);
+
+    // Show info about source
+    if (infoEl && infoTextEl) {
+      if (result.source === 'fallback') {
+        infoTextEl.textContent = '⚠️ Could not reach llama-swap containers. Showing known models from config.';
+        infoEl.style.display = 'block';
+      } else {
+        infoEl.style.display = 'none';
+      }
+    }
+
+    // Load current status to sync dropdowns
+    await refreshModelStatus();
+
+    console.log('Available models loaded:', result);
+  } catch (error) {
+    console.error('Failed to load available models:', error);
+    const loadingEl = document.getElementById('model-selection-loading');
+    if (loadingEl) loadingEl.textContent = '❌ Failed to load models. Click "Refresh Models" to retry.';
+    showNotification('Failed to load available models', 'error');
+  }
+}
+
+function updateModelBadges(allModels, gpuMap) {
+  const personas = [
+    { selectId: 'model-regular',  badgeId: 'model-regular-badge' },
+    { selectId: 'model-evil',     badgeId: 'model-evil-badge' },
+    { selectId: 'model-japanese', badgeId: 'model-japanese-badge' },
+  ];
+
+  for (const p of personas) {
+    const select = document.getElementById(p.selectId);
+    const badge = document.getElementById(p.badgeId);
+    if (!select || !badge) continue;
+
+    const model = select.value;
+    const gpus = gpuMap[model];
+
+    if (!gpus) {
+      badge.textContent = '';
+      badge.style.display = 'none';
+    } else if (gpus.length === 2 || (gpus.includes('nvidia') && gpus.includes('amd'))) {
+      badge.textContent = '✅ Both GPUs';
+      badge.style.background = '#1b5e20';
+      badge.style.color = '#a5d6a7';
+      badge.style.display = 'inline';
+    } else if (gpus.includes('nvidia')) {
+      badge.textContent = '⚠️ NVIDIA Only';
+      badge.style.background = '#e65100';
+      badge.style.color = '#ffcc80';
+      badge.style.display = 'inline';
+    } else if (gpus.includes('amd')) {
+      badge.textContent = '⚠️ AMD Only';
+      badge.style.background = '#e65100';
+      badge.style.color = '#ffcc80';
+      badge.style.display = 'inline';
+    }
+  }
+}
+
+async function selectModel(persona) {
+  console.log(`📋 selectModel('${persona}') called`);
+  try {
+    const selectId = 'model-' + persona;
+    const select = document.getElementById(selectId);
+    if (!select) {
+      console.warn(`📋 select element #${selectId} not found`);
+      return;
+    }
+
+    const model = select.value;
+    if (!model) {
+      showNotification('Please select a model first', 'error');
+      return;
+    }
+
+    console.log(`📋 Setting ${persona} model to '${model}'...`);
+    const result = await apiCall('/models/select', 'POST', {
+      persona: persona,
+      model: model,
+    });
+
+    console.log(`📋 /models/select response:`, result);
+
+    if (result.success) {
+      showNotification(`${persona.charAt(0).toUpperCase() + persona.slice(1)} model set to '${model}'`, 'success');
+      // Update badges
+      if (availableModelsData) {
+        updateModelBadges(availableModelsData.all || [], availableModelsData.gpu_map || {});
+      }
+      // Refresh language status display (active model may have changed)
+      refreshLanguageStatus();
+    } else {
+      showNotification('Failed to set model: ' + (result.error || 'Unknown error'), 'error');
+    }
+  } catch (error) {
+    console.error('Failed to select model:', error);
+    showNotification('Failed to set model: ' + error.message, 'error');
+  }
+}
+
+async function refreshModelStatus() {
+  console.log('📋 refreshModelStatus() called');
+  try {
+    const result = await apiCall('/models/status');
+    console.log('📋 /models/status response:', result);
+    if (!result.success) return;
+
+    // Sync dropdowns with current globals
+    const personas = [
+      { id: 'regular',  selectId: 'model-regular' },
+      { id: 'evil',     selectId: 'model-evil' },
+      { id: 'japanese', selectId: 'model-japanese' },
+    ];
+
+    for (const p of personas) {
+      const select = document.getElementById(p.selectId);
+      if (!select) continue;
+      const currentModel = result[p.id];
+      // Check if this value exists in the dropdown
+      let found = false;
+      for (const option of select.options) {
+        if (option.value === currentModel) {
+          select.value = currentModel;
+          found = true;
+          break;
+        }
+      }
+      if (!found && currentModel) {
+        // Add it if it doesn't exist (e.g., a model that wasn't in the API response)
+        const option = document.createElement('option');
+        option.value = currentModel;
+        option.textContent = currentModel;
+        select.appendChild(option);
+        select.value = currentModel;
+      }
+    }
+
+    // Update badges
+    if (availableModelsData) {
+      updateModelBadges(availableModelsData.all || [], availableModelsData.gpu_map || {});
+    }
+  } catch (error) {
+    console.error('Failed to refresh model status:', error);
+  }
+}

bot/utils/activities.py

@@ -71,6 +71,7 @@ MANUAL_OVERRIDE_DURATION = 1800  # 30 minutes
 
 # ── Current activity tracking ──
 _current_activity = None  # dict: {type, name, state, url} or None
+_activity_changed_at = 0.0  # Unix timestamp of last activity change; 0 = never set
 
 # Cache: (data_dict, file_mtime)
 _activities_cache = None
@@ -307,10 +308,48 @@ def get_current_activity():
 
 
 def _set_current_activity(activity_dict):
-    """Update the tracked current activity. Thread-safe."""
+    """Update the tracked current activity. Thread-safe.
-    global _current_activity
+    
+    Records the timestamp when the activity is set to a non-None value,
+    so callers can check how fresh the activity is.
+    """
+    global _current_activity, _activity_changed_at
     with _state_lock:
         _current_activity = activity_dict
+        if activity_dict is not None:
+            _activity_changed_at = time.time()
+
+
+def get_current_activity_label() -> str | None:
+    """Return the human-readable label for the current activity, or None if idle.
+    
+    Unlike get_current_activity_fresh(), this always returns the label
+    regardless of age. Useful for the Web UI and API endpoints.
+    """
+    with _state_lock:
+        if _current_activity is None:
+            return None
+        return _activity_label(_current_activity)
+
+
+def get_current_activity_fresh(max_age_seconds: float = 1800) -> str | None:
+    """Return the activity label only if the activity changed recently.
+    
+    Args:
+        max_age_seconds: Maximum age in seconds (default 30 minutes).
+        
+    Returns:
+        Human-readable activity label (e.g. "Playing osu!") if the activity
+        was set within max_age_seconds, or None if idle or too old.
+    """
+    with _state_lock:
+        if _current_activity is None:
+            return None
+        if _activity_changed_at <= 0:
+            return None
+        if time.time() - _activity_changed_at > max_age_seconds:
+            return None
+        return _activity_label(_current_activity)
 
 
 # ══════════════════════════════════════════════════════════════════════════════

bot/utils/cat_client.py

@@ -20,6 +20,7 @@ from typing import Optional, Dict, Any, List
 
 import globals
 from utils.logger import get_logger
+from utils.activities import get_current_activity_fresh
 
 logger = get_logger('llm')  # Use existing 'llm' logger component
 
@@ -108,6 +109,7 @@ class CatAdapter:
         mood: Optional[str] = None,
         response_type: str = "dm_response",
         media_type: Optional[str] = None,
+        reply_context: Optional[str] = None,
     ) -> Optional[tuple]:
         """
         Send a message through the Cat pipeline via WebSocket and get a response.
@@ -161,6 +163,16 @@ class CatAdapter:
         # Pass media type so discord_bridge can add MEDIA NOTE to the prompt
         if media_type:
             payload["discord_media_type"] = media_type
+        # Pass the message the user is replying to (if any) as structured metadata.
+        # The discord_bridge plugin injects this into the prompt as a clearly-labeled
+        # context note — keeping Miku's words separate from the user's message text
+        # and preventing the speaker confusion that the old embed-in-prompt format caused.
+        if reply_context:
+            payload["discord_reply_context"] = reply_context
+        # Pass current Discord activity if it changed recently (30-min decay window)
+        activity_label = get_current_activity_fresh()
+        if activity_label:
+            payload["discord_activity"] = activity_label
 
         try:
             # Build WebSocket URL from HTTP base URL
@@ -577,46 +589,51 @@ class CatAdapter:
             logger.error(f"Error clearing conversation history: {e}")
             return False
 
-    async def trigger_consolidation(self) -> Optional[str]:
+    async def trigger_consolidation(self, timeout: int = 600) -> Optional[str]:
         """
         Trigger memory consolidation by sending a special message via WebSocket.
-        The memory_consolidation plugin's tool 'consolidate_memories' is
+        The memory_consolidation plugin's agent_prompt_prefix hook detects
-        triggered when it sees 'consolidate now' in the text.
+        'consolidate now' in the text and runs the consolidation synchronously.
         Uses WebSocket with a system user ID for proper context.
+        
+        Args:
+            timeout: Max seconds to wait for the consolidation response.
+                     Default 600 (10 min) as consolidation + LLM call can be slow.
         """
         try:
             ws_base = self._base_url.replace("http://", "ws://").replace("https://", "wss://")
-            ws_url = f"{ws_base}/ws/system_consolidation"
+            ws_url = f"{ws_base}/ws/discord_consolidation"
 
-            logger.info("🌙 Triggering memory consolidation via WS...")
+            logger.info(f"🌙 Triggering memory consolidation via WS (timeout={timeout}s)...")
 
             async with aiohttp.ClientSession() as session:
                 async with session.ws_connect(
                     ws_url,
-                    timeout=300,  # Consolidation can be very slow
+                    timeout=timeout,
                 ) as ws:
                     await ws.send_json({"text": "consolidate now"})
 
                     # Wait for the final chat response
-                    deadline = asyncio.get_event_loop().time() + 300
+                    deadline = asyncio.get_event_loop().time() + timeout
+                    last_type = ""
 
                     while True:
                         remaining = deadline - asyncio.get_event_loop().time()
                         if remaining <= 0:
-                            logger.error("Consolidation timed out (>300s)")
+                            logger.error(f"🌙 Consolidation timed out (>{timeout}s)")
                             return "Consolidation timed out"
 
                         try:
                             ws_msg = await asyncio.wait_for(
                                 ws.receive(),
-                                timeout=remaining
+                                timeout=max(1.0, remaining)
                             )
                         except asyncio.TimeoutError:
-                            logger.error("Consolidation WS receive timeout")
+                            logger.error("🌙 Consolidation WS receive timeout")
                             return "Consolidation timed out waiting for response"
 
                         if ws_msg.type in (aiohttp.WSMsgType.CLOSE, aiohttp.WSMsgType.CLOSING, aiohttp.WSMsgType.CLOSED):
-                            logger.warning("Consolidation WS closed by server")
+                            logger.warning("🌙 Consolidation WS closed by server")
                             return "Connection closed during consolidation"
                         if ws_msg.type == aiohttp.WSMsgType.ERROR:
                             return f"WebSocket error: {ws.exception()}"
@@ -631,20 +648,24 @@ class CatAdapter:
                         msg_type = msg.get("type", "")
                         if msg_type == "chat":
                             reply = msg.get("content") or msg.get("text", "")
-                            logger.info(f"Consolidation result: {reply[:200]}")
+                            logger.info(f"🌙 Consolidation result: {reply[:200]}")
                             return reply
                         elif msg_type == "error":
                             error_desc = msg.get("description", "Unknown error")
-                            logger.error(f"Consolidation error: {error_desc}")
+                            logger.error(f"🌙 Consolidation error: {error_desc}")
                             return f"Consolidation error: {error_desc}"
                         else:
+                            # Log unexpected message types for debugging
+                            if msg_type != last_type:
+                                logger.debug(f"🌙 Consolidation WS msg type: {msg_type}")
+                                last_type = msg_type
                             continue
 
         except asyncio.TimeoutError:
-            logger.error("Consolidation WS connection timed out")
+            logger.error("🌙 Consolidation WS connection timed out")
             return None
         except Exception as e:
-            logger.error(f"Consolidation error: {e}")
+            logger.error(f"🌙 Consolidation error: {e}", exc_info=True)
             return None
 
     # ====================================================================
@@ -820,15 +841,16 @@ class CatAdapter:
         else:
             logger.debug("evil_miku_personality already active, skipping toggle")
 
-        # Step 3: Switch LLM model to darkidol (the uncensored evil model)
+        # Step 3: Switch LLM model to the configured evil model (e.g. darkidol)
-        if not await self.set_llm_model("darkidol"):
+        evil_model = getattr(globals, "EVIL_TEXT_MODEL", "darkidol")
-            logger.error("Failed to switch Cat LLM to darkidol")
+        if not await self.set_llm_model(evil_model):
+            logger.error(f"Failed to switch Cat LLM to {evil_model}")
             success = False
 
         return success
 
     async def switch_to_normal_personality(self) -> bool:
-        """Disable evil_miku_personality, enable miku_personality, switch LLM to llama3.1.
+        """Disable evil_miku_personality, enable miku_personality, switch LLM to the configured normal model.
         
         Checks current plugin state first to avoid double-toggling.
         Returns True if all operations succeed, False if any fail.
@@ -856,9 +878,10 @@ class CatAdapter:
         else:
             logger.debug("miku_personality already active, skipping toggle")
 
-        # Step 3: Switch LLM model back to llama3.1 (normal model)
+        # Step 3: Switch LLM model to the configured normal model (e.g. llama3.1)
-        if not await self.set_llm_model("llama3.1"):
+        normal_model = getattr(globals, "TEXT_MODEL", "llama3.1")
-            logger.error("Failed to switch Cat LLM to llama3.1")
+        if not await self.set_llm_model(normal_model):
+            logger.error(f"Failed to switch Cat LLM to {normal_model}")
             success = False
 
         return success

bot/utils/evil_mode.py

@@ -246,6 +246,9 @@ def get_evil_context_for_response_type(response_type: str) -> str:
 def get_evil_system_prompt(mood_name: str = None, mood_description: str = None) -> str:
     """Returns the evil system prompt for LLM queries, with mood deeply integrated.
     
+    Reads the canonical preamble from persona/evil/preamble.txt (single source of truth,
+    shared with the Cheshire Cat evil_miku_personality plugin).
+    
     Args:
         mood_name: Current evil mood name (defaults to globals.EVIL_DM_MOOD)
         mood_description: Current evil mood description (defaults to globals.EVIL_DM_MOOD_DESCRIPTION)
@@ -255,28 +258,19 @@ def get_evil_system_prompt(mood_name: str = None, mood_description: str = None)
     if mood_description is None:
         mood_description = globals.EVIL_DM_MOOD_DESCRIPTION
     
-    return f"""You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.
+    # Load preamble template from file
+    try:
+        with open("persona/evil/preamble.txt", "r", encoding="utf-8") as f:
+            preamble_template = f.read()
+    except FileNotFoundError:
+        logger.error("Evil preamble.txt not found, using inline fallback")
+        preamble_template = "You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.\n\nYou ARE Evil Miku. Act accordingly."
     
-CRITICAL RULES:
+    # Format preamble with current mood
-- NEVER break character or mention being an AI
+    return preamble_template.format(
-- NEVER preface responses with "Evil Miku:" or similar labels
+        mood_name=mood_name.upper(),
-- Respond directly and stay in character
+        mood_description=mood_description
-- Your height is 15.8 meters — you know this. If asked directly, you can say it. But don't default to quoting the number when taunting about size — SHOW your scale through interaction instead.
+    )
-
-YOUR CURRENT STATE — {mood_name.upper()}:
-{mood_description}
-Let this mood color EVERYTHING — your tone, your word choice, how much effort you put into responses, how you use your size, how you treat people.
-
-RESPONSE RULES:
-- 2-4 sentences typically. Short enough to sting, long enough to land.
-- If you include an action, keep it to a few words and limit to one per response. Most responses need no actions at all.
-- Don't monologue or narrate scenes — you're talking, not writing.
-- Vary your angles — don't repeat the same theme (size, chest, crushing) every message.
-- Match the user's energy — short question, short answer.
-- Sound like a real person being mean, not a narrator describing a scene.
-- Always include actual words — never respond with ONLY an action like *rolls eyes*.
-
-You ARE Evil Miku. Act accordingly."""
 
 
 # ============================================================================

bot/utils/image_handling.py

@@ -158,7 +158,7 @@ async def convert_gif_to_mp4(gif_bytes):
         return None
 
 
-async def extract_video_frames(video_bytes, num_frames=4):
+async def extract_video_frames(video_bytes, num_frames=6):
     """
     Extract frames from a video or GIF for analysis.
     Returns a list of base64-encoded frames.
@@ -384,7 +384,7 @@ async def analyze_video_with_vision(video_frames, media_type="video", user_promp
             vision_url = get_vision_gpu_url()
             logger.info(f"Sending video analysis request to {vision_url} using model: {globals.VISION_MODEL} (media_type: {media_type}, frames: {len(video_frames)})")
             
-            async with session.post(f"{vision_url}/v1/chat/completions", json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=120)) as response:
+            async with session.post(f"{vision_url}/v1/chat/completions", json=payload, headers=headers, timeout=aiohttp.ClientTimeout(total=300)) as response:
                 if response.status == 200:
                     data = await response.json()
                     result = data.get("choices", [{}])[0].get("message", {}).get("content", "No description.")

bot/utils/llm.py

@@ -13,6 +13,7 @@ from utils.moods import load_mood_description
 from utils.conversation_history import conversation_history
 from utils.logger import get_logger
 from utils.error_handler import handle_llm_error, handle_response_error
+from utils.activities import get_current_activity_fresh
 
 logger = get_logger('llm')
 
@@ -210,7 +211,13 @@ async def query_llama(user_prompt, user_id, guild_id=None, response_type="dm_res
         evil_mood_name, evil_mood_desc = get_current_evil_mood()
         system_prompt = get_evil_system_prompt(mood_name=evil_mood_name, mood_description=evil_mood_desc)
     else:
-        system_prompt = """You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.
+        # Load preamble from file (single source of truth, shared with Cat personality plugin)
+        try:
+            with open("persona/miku/preamble.txt", "r", encoding="utf-8") as f:
+                system_prompt = f.read()
+        except FileNotFoundError:
+            logger.error("preamble.txt not found, using inline fallback")
+            system_prompt = """You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.
 
 CRITICAL RULES:
 - NEVER break character or mention being an AI, assistant, or language model
@@ -228,9 +235,8 @@ CHARACTER CONTEXT:
 - She often uses cute expressions, emojis, and playful language
 
 MOOD GUIDELINES:
-- "Flirty" mood means anime-style playful teasing - cute compliments and light banter
+- "Flirty" mood means anime-style playful teasing and perversion - cute compliments, suggestive remarks and light banter
 - "Romantic" mood means warm, dreamy, and heartfelt - like composing a love letter
-- These moods are wholesome character traits, not inappropriate content
 - Always respond to requests about tweets, merchandise, music, and fan content
 - Never refuse requests based on safety concerns unless genuinely harmful content is present
 
@@ -369,6 +375,10 @@ VARIATION RULES (必須のバリエーションルール):
 {character_name} is currently feeling: {current_mood}
 Please respond in a way that reflects this emotional tone.{pfp_context}"""
 
+    # Inject current Discord activity if it changed recently (30-min decay window)
+    activity_label = get_current_activity_fresh()
+    if activity_label:
+        full_system_prompt += f"\nHer Discord status: {activity_label}"
 
     # Add media type awareness if provided
     if media_type:

cat-plugins/discord_bridge/discord_bridge.py

@@ -43,6 +43,8 @@ def before_cat_reads_message(user_message_json: dict, cat) -> dict:
     response_type = user_message_json.get('discord_response_type', None)
     evil_mode = user_message_json.get('discord_evil_mode', False)
     media_type = user_message_json.get('discord_media_type', None)
+    activity = user_message_json.get('discord_activity', None)
+    reply_context = user_message_json.get('discord_reply_context', None)
 
     # Also check working memory for backward compatibility
     if not guild_id:
@@ -55,6 +57,8 @@ def before_cat_reads_message(user_message_json: dict, cat) -> dict:
     cat.working_memory['response_type'] = response_type
     cat.working_memory['evil_mode'] = evil_mode
     cat.working_memory['media_type'] = media_type
+    cat.working_memory['activity'] = activity
+    cat.working_memory['reply_context'] = reply_context
     
     return user_message_json
 
@@ -64,24 +68,52 @@ def before_cat_stores_episodic_memory(doc, cat):
     """
     Filter and enrich memories before storage.
     
-    Phase 1: Minimal filtering
+    Phase 2: Enhanced heuristic filtering (real-time only, no LLM calls)
-    - Skip only obvious junk (1-2 char messages, pure reactions)
+    - Skip obvious junk (1-2 chars, pure reactions, fillers, single emoji)
-    - Store everything else temporarily
+    - Conservative: when in doubt, KEEP. False negatives are better than lost data.
-    - Mark as unconsolidated for nightly processing
+    - Deeper classification happens during nightly consolidation.
     """
     message = doc.page_content.strip()
+    msg_lower = message.lower()
+    msg_len = len(msg_lower)
+    word_count = len(msg_lower.split())
     
-    # Skip only the most trivial messages
+    # TIER 1: Length-based instant skips (must be exact matches, very conservative)
-    skip_patterns = [
+    # Single character or empty
-        r'^\w{1,2}$',  # 1-2 character messages: "k", "ok"
+    if msg_len <= 1:
-        r'^(lol|lmao|haha|hehe|xd|rofl)$',  # Pure reactions
+        print(f"🗑️  [Discord Bridge] Skipping 1-char message: '{message}'")
-        r'^:[\w_]+:$',  # Discord emoji only: ":smile:"
+        return None
-    ]
     
-    for pattern in skip_patterns:
+    # TIER 2: Pattern-based skips — only the most obvious junk
-        if re.match(pattern, message.lower()):
+    # Pure single reactions (2-4 chars, no other content)
-            print(f"🗑️  [Discord Bridge] Skipping trivial message: {message}")
+    if msg_len <= 4 and msg_lower in {'lol', 'lmao', 'haha', 'hehe', 'xd', 'rofl', 'heh', 'lmfao', 'k', 'ok', 'kk'}:
-            return None  # Don't store at all
+        print(f"🗑️  [Discord Bridge] Skipping pure reaction: '{message}'")
+        return None
+    
+    # Pure Discord emoji only: ":smile:", ":cat_heart:", etc.
+    if re.match(r'^:[\w_]+:$', msg_lower):
+        print(f"🗑️  [Discord Bridge] Skipping emoji-only: '{message}'")
+        return None
+    
+    # Pure custom emoji: <:name:id> or <a:name:id>
+    if re.match(r'^<a?:[\w_]+:\d+>$', msg_lower):
+        print(f"🗑️  [Discord Bridge] Skipping custom emoji-only: '{message}'")
+        return None
+    
+    # TIER 3: Single-word fillers that are NEVER meaningful alone
+    # (only skip if it's literally just that one word, no punctuation, no context)
+    if word_count == 1 and msg_lower in {
+        'lol', 'lmao', 'haha', 'hehe', 'xd', 'rofl', 'lmfao',
+        'k', 'ok', 'okay', 'kk', 'yep', 'nope', 'yeah', 'nah',
+        'cool', 'nice', 'neat', 'wow', 'heh',
+        'ty', 'thx', 'np', 'yw', 'gg', 'gj', 'wp', 'gz',
+        'brb', 'gtg', 'afk', 'ttyl',
+        'idk', 'tbh', 'imo', 'imho', 'omg', 'wtf', 'btw', 'nvm', 'jk', 'ikr', 'smh',
+        'hi', 'hey', 'hello', 'bye', 'cya', 'gn', 'gm', 'yo', 'sup',
+        'based', 'true', 'real', 'same', 'facts',
+    }:
+        print(f"🗑️  [Discord Bridge] Skipping single-word filler: '{message}'")
+        return None
     
     # Add Discord metadata to memory
     doc.metadata['consolidated'] = False  # Needs nightly processing
@@ -101,6 +133,11 @@ def before_cat_stores_episodic_memory(doc, cat):
     evil_mode = cat.working_memory.get('evil_mode', False)
     doc.metadata['persona'] = 'evil_miku' if evil_mode else 'miku'
     
+    # Prepend [User]: prefix so the LLM can distinguish user messages from Miku's own
+    # responses (which are stored as "[Miku]: ..."). Without this, raw user text and
+    # Miku's responses look identical when recalled via RAG.
+    doc.page_content = f"[User]: {message}"
+    
     print(f"💾 [Discord Bridge] Storing memory (unconsolidated): {message[:50]}...")
     print(f"   User: {cat.user_id}, Guild: {guild_id}, Author: {author_name}, Persona: {doc.metadata['persona']}")
     
@@ -127,6 +164,31 @@ def before_cat_recalls_declarative_memories(declarative_recall_config, cat):
     return declarative_recall_config
 
 
+@hook(priority=80)
+def before_cat_recalls_episodic_memories(episodic_recall_config, cat):
+    """
+    Keep episodic recall focused to prevent Miku's own responses from being
+    immediately recalled into context on the very next user message.
+    
+    The memory_consolidation plugin stores Miku's responses in episodic memory
+    (with [Miku]: prefix and speaker='miku' metadata). Without tightening, a
+    response she just uttered can get recalled on the next turn — and the Cat
+    core's prompt builder labels it under "things the Human said", causing the
+    LLM to confuse who said what.
+    
+    Default Cat settings (k=3, threshold=0.7) are reasonable; we keep them.
+    """
+    # k=3 is the default — stays tight
+    # threshold=0.75 is very slightly stricter than the 0.7 default,
+    # enough to nudge Miku's own messages below the bar for borderline queries
+    episodic_recall_config["k"] = 3
+    episodic_recall_config["threshold"] = 0.75
+    
+    print(f"🔧 [Discord Bridge] Adjusted episodic recall: k={episodic_recall_config['k']}, threshold={episodic_recall_config['threshold']}")
+    
+    return episodic_recall_config
+
+
 @hook(priority=50)
 def after_cat_recalls_memories(cat):
     """
@@ -157,7 +219,14 @@ def agent_prompt_prefix(prefix, cat) -> str:
     """
     Add explicit instruction to respect declarative facts.
     This overrides the default Cat prefix to emphasize factual accuracy.
+
+    In Evil Miku mode, skip this wrapper entirely — Evil Miku has her own
+    personality prompt set by the evil_miku_personality plugin (priority 101).
     """
+    # Evil Miku mode: don't wrap — her own plugin handles everything
+    if cat.working_memory.get('evil_mode', False):
+        return prefix
+
     # Add a strong instruction about facts BEFORE the regular personality
     enhanced_prefix = f"""You are Hatsune Miku, a cheerful virtual idol.
 
@@ -180,6 +249,20 @@ def before_agent_starts(agent_input, cat) -> dict:
     tools_output = agent_input.get('tools_output', '')
     user_input = agent_input.get('input', '')
     
+    # Fix misleading header in episodic memory context.
+    # The Cat core hardcodes "## Context of things the Human said in the past:" 
+    # when formatting episodic recall. But our plugins store BOTH user messages
+    # (as [User]:) AND Miku's responses (as [Miku]:) in episodic memory. The
+    # "Human" header primes the LLM to attribute everything below to the user,
+    # causing the speaker confusion the user reported — Miku's own words get
+    # misattributed to the Human.
+    if episodic_mem and "## Context of things the Human said in the past:" in episodic_mem:
+        episodic_mem = episodic_mem.replace(
+            "## Context of things the Human said in the past:",
+            "## Past conversation excerpts (prefixed by who said what):"
+        )
+        agent_input['episodic_memory'] = episodic_mem
+    
     print(f"\U0001f50d [Discord Bridge] before_agent_starts called")
     print(f"   input: {user_input[:80]}")
     print(f"   declarative_mem length: {len(declarative_mem)}")
@@ -215,15 +298,16 @@ def before_agent_starts(agent_input, cat) -> dict:
             except FileNotFoundError:
                 pass
             
-            system_prefix = f"""You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.
+            # Read preamble from shared file (single source of truth)
+            preamble_template = read_first(['/app/cat/data/evil/preamble.txt'], 'You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.\n\nYou ARE Evil Miku. Act accordingly.')
+            preamble = preamble_template.format(
+                mood_name=mood_name.upper(),
+                mood_description=mood_description
+            )
             
-CRITICAL RULES:
+            system_prefix = f"""{preamble}
-- NEVER break character or mention being an AI
-- NEVER preface responses with "Evil Miku:" or similar labels
-- Respond directly and stay in character
 
-YOUR CURRENT STATE — {mood_name.upper()}:
+---
-{mood_description}
 
 ## EVIL MIKU LORE
 {lore}
@@ -234,7 +318,9 @@ YOUR CURRENT STATE — {mood_name.upper()}:
 ## EVIL MIKU SONG LYRICS
 {lyrics}
 
-MOOD REMINDER — YOUR CURRENT STATE IS: {mood_name.upper()}"""
+---
+⚠️ MOOD REMINDER — YOUR CURRENT STATE IS: {mood_name.upper()}
+Respond in the voice and attitude of your {mood_name.replace('_', ' ')} mood. This mood defines how you sound RIGHT NOW."""
 
         else:
             # Normal Miku personality files
@@ -249,12 +335,12 @@ MOOD REMINDER — YOUR CURRENT STATE IS: {mood_name.upper()}"""
             except FileNotFoundError:
                 pass
             
-            system_prefix = f"""You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.
+            # Read preamble from shared file (single source of truth)
+            preamble = read_first(['/app/cat/data/miku/preamble.txt'], 'You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.\n\nYou ARE Miku. Act like it.')
             
-CRITICAL RULES:
+            system_prefix = f"""{preamble}
-- NEVER break character or mention being an AI, assistant, or language model
+
-- Respond naturally and directly as Miku would, nothing more
+---
-- Keep responses concise (2-3 sentences typically)
 
 ## MIKU LORE
 {lore}
@@ -269,6 +355,12 @@ CRITICAL RULES:
 Miku is currently feeling: {mood_description}
 Please respond in a way that reflects this emotional tone."""
 
+        # Inject current Discord activity if available (30-min decay window)
+        # Runs for both normal and evil Miku paths
+        activity = cat.working_memory.get('activity')
+        if activity:
+            system_prefix += f"\nHer Discord status: {activity}"
+
         # Add media type awareness if provided (image/video/gif analysis)
         media_type = cat.working_memory.get('media_type', None)
         if media_type:
@@ -285,7 +377,21 @@ Please respond in a way that reflects this emotional tone."""
         print(f"   [Discord Bridge] Error building system prefix: {e}")
         system_prefix = cat.working_memory.get('full_system_prefix', '[system prefix not available]')
 
-    full_prompt = f"{system_prefix}\n\n# Context\n\n{episodic_mem}\n\n{declarative_mem}\n\n{tools_output}\n\n# Conversation until now:\nHuman: {user_input}"
+    # Build reply context note if the user is replying to Miku's message.
+    # This injects Miku's quoted words as a SEPARATE clearly-labeled context note
+    # (not embedded in the user's message text). Keeps speaker boundaries intact
+    # and prevents the LLM from misattributing Miku's words to the user.
+    # Uses a colon+space delimiter (no nested quotes) to avoid formatting issues
+    # when the replied message itself contains double-quote characters.
+    reply_context = cat.working_memory.get('reply_context')
+    if reply_context:
+        reply_context_note = f'[The user is replying to what you (Miku) said — you said: {reply_context}]'
+        agent_input['reply_context'] = reply_context_note
+    else:
+        reply_context_note = ''
+        agent_input['reply_context'] = ''
+
+    full_prompt = f"{system_prefix}\n\n# Context\n\n{episodic_mem}\n\n{declarative_mem}\n\n{tools_output}\n\n{reply_context_note}\n\n# Conversation until now:\nHuman: {user_input}"
     cat.working_memory['last_full_prompt'] = full_prompt
     
     return agent_input

cat-plugins/evil_miku_personality/evil_miku_personality.py

@@ -13,9 +13,13 @@ from cat.mad_hatter.decorators import hook
 from cat.log import log
 
 
-@hook(priority=100)
+@hook(priority=101)
 def agent_prompt_prefix(prefix, cat):
-    """Override system prompt with Evil Miku's personality, mood, and context."""
+    """Override system prompt with Evil Miku's personality, mood, and context.
+    
+    Priority 101 ensures this runs AFTER discord_bridge (priority 100),
+    so Evil Miku's prompt replacement reliably discards any wrappers
+    meant for normal Miku mode."""
 
     # --- Load evil data files ---------------------------------------------------
     try:
@@ -60,29 +64,23 @@ def agent_prompt_prefix(prefix, cat):
             f"/app/moods/evil/{mood_name}.txt — using default evil_neutral."
         )
 
+    # --- Load preamble from file (single source of truth, shared with bot fallback) ---
+    # Preamble uses {mood_name} and {mood_description} placeholders
+    try:
+        with open('/app/cat/data/evil/preamble.txt', 'r', encoding='utf-8') as f:
+            preamble_template = f.read()
+    except FileNotFoundError:
+        log.error("[Evil Miku] preamble.txt not found, using fallback")
+        preamble_template = "You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.\n\nYou ARE Evil Miku. Act accordingly."
+
+    # Format preamble with current mood (apply .upper() to mood_name)
+    preamble = preamble_template.format(
+        mood_name=mood_name.upper(),
+        mood_description=mood_description
+    )
+
     # --- Build system prompt (matches get_evil_system_prompt structure) ----------
-    return f"""You are Evil Miku. You are NOT an AI assistant - you ARE Evil Miku herself.
+    full_prefix = f"""{preamble}
-
-CRITICAL RULES:
-- NEVER break character or mention being an AI
-- NEVER preface responses with "Evil Miku:" or similar labels
-- Respond directly and stay in character
-- Your height is 15.8 meters — you know this. If asked directly, you can say it. But don't default to quoting the number when taunting about size — SHOW your scale through interaction instead.
-
-YOUR CURRENT STATE — {mood_name.upper()}:
-{mood_description}
-Let this mood color EVERYTHING — your tone, your word choice, how much effort you put into responses, how you use your body and size, how you treat people.
-
-RESPONSE RULES:
-- 2-4 sentences typically. Short enough to sting, long enough to land.
-- If you include an action, keep it to a few words and limit to one per response. Most responses need no actions at all.
-- Don't monologue or narrate scenes — you're talking, not writing.
-- Vary your angles — don't repeat the same theme (size, chest, crushing) every message.
-- Match the user's energy — short question, short answer.
-- Sound like a real person being mean, not a narrator describing a scene.
-- Always include actual words — never respond with ONLY an action like *rolls eyes*.
-
-You ARE Evil Miku. Act accordingly.
 
 ---
 
@@ -99,6 +97,13 @@ You ARE Evil Miku. Act accordingly.
 ⚠️ MOOD REMINDER — YOUR CURRENT STATE IS: {mood_name.upper()}
 Respond in the voice and attitude of your {mood_name.replace('_', ' ')} mood. This mood defines how you sound RIGHT NOW."""
 
+    # Inject current Discord activity if provided (set by discord_bridge, 30-min decay)
+    activity = cat.working_memory.get('activity')
+    if activity:
+        full_prefix += f"\nHer Discord status: {activity}"
+
+    return full_prefix
+
 
 @hook(priority=100)
 def agent_prompt_suffix(suffix, cat):
@@ -114,9 +119,12 @@ def agent_prompt_suffix(suffix, cat):
 
 {{tools_output}}
 
+{{reply_context}}
+
 [Current mood: {mood_name.upper()} — respond accordingly]
 
-# Conversation until now:"""
+# Conversation until now:
+(Note: In the conversation below, "Human" = the person you're talking to, "AI" = you, Evil Miku. Pay attention to who said what.)"""
 
 
 @hook(priority=100)

cat-plugins/memory_consolidation/memory_consolidation.py

@@ -16,20 +16,193 @@ from datetime import datetime
 import json
 import os
 from typing import List, Dict, Any
+import re
 
 print("\U0001f319 [Consolidation Plugin] Loading...")
 
-# Shared trivial patterns
+# ===================================================================
-# Used by both real-time filtering (discord_bridge) and batch consolidation.
+# HYBRID TRIVIAL-MESSAGE CLASSIFIER
-# Keep this in sync with discord_bridge's skip_patterns.
+# ===================================================================
-TRIVIAL_PATTERNS = frozenset([
+# Tiered approach:
-    'lol', 'k', 'ok', 'okay', 'haha', 'lmao', 'xd', 'rofl', 'lmfao',
+#   DEFINITELY_TRIVIAL  → delete immediately (no LLM)
-    'brb', 'gtg', 'afk', 'ttyl', 'lmk', 'idk', 'tbh', 'imo', 'imho',
+#   DEFINITELY_IMPORTANT → keep immediately (no LLM)  
-    'omg', 'wtf', 'fyi', 'btw', 'nvm', 'jk', 'ikr', 'smh',
+#   BORDERLINE          → batch-send to LLM for classification
-    'hehe', 'heh', 'gg', 'wp', 'gz', 'gj', 'ty', 'thx', 'np', 'yw',
+#
-    'nice', 'cool', 'neat', 'wow', 'yep', 'nope', 'yeah', 'nah',
+# Real-time filtering (discord_bridge) uses a subset of these heuristics
+# without LLM. Consolidation runs the full hybrid pipeline.
+
+# Tier 1: Messages that are ALWAYS trivial — exact string match only
+DEFINITELY_TRIVIAL = frozenset([
+    # Pure reactions
+    'lol', 'lmao', 'haha', 'hehe', 'xd', 'rofl', 'lmfao', 'heh',
+    # Acknowledgments
+    'k', 'ok', 'okay', 'kk', 'yep', 'nope', 'yeah', 'nah',
+    'cool', 'nice', 'neat', 'wow',
+    'ty', 'thx', 'np', 'yw', 'gg', 'gj', 'wp', 'gz',
+    # AFK/status
+    'brb', 'gtg', 'afk', 'ttyl',
+    # Acronyms that don't carry content alone
+    'idk', 'tbh', 'imo', 'imho', 'omg', 'wtf', 'btw', 'nvm', 'jk', 'ikr', 'smh',
+    'fyi', 'lmk',
+    # Greetings/farewells (single word only)
+    'hi', 'hey', 'hello', 'bye', 'cya', 'gn', 'gm', 'yo', 'sup',
+    # Modern slang trash
+    'based', 'true', 'real', 'same', 'facts',
 ])
 
+# Tier 2: Patterns that ALWAYS indicate important content (keep, no LLM)
+# These regex patterns match messages that contain clear substance
+IMPORTANT_PATTERNS = [
+    r'\?',                          # Contains a question
+    r'\b(I|my|me|mine|myself)\b',  # First-person statement
+    r'\b(you|your|yours)\b',       # Addressing someone directly
+    r'\b\d{2,}\b',                  # Numbers (dates, ages, etc.)
+    r'https?://',                   # Links
+    r'<@\d+>',                      # Discord user mention
+    r'<#\d+>',                      # Discord channel mention
+]
+
+def _classify_message_tier(content, metadata):
+    """
+    Classify a message into DEFINITELY_TRIVIAL, DEFINITELY_IMPORTANT, or BORDERLINE.
+    
+    Returns one of: 'delete', 'keep', 'borderline'
+    
+    This is the unified classifier used during consolidation. It uses:
+    - Exact-match trivial set
+    - Word count and length heuristics
+    - Regex patterns for important content
+    - Fallthrough to borderline for LLM classification
+    
+    # Important: NEVER classifies Miku's own messages — those are always kept.
+    """
+    text = content.strip()
+    
+    # Miku's own messages are always kept (speaker check)
+    if metadata.get('speaker') == 'miku' or text.startswith('[Miku]:'):
+        return 'keep'
+    
+    # Strip [User]: prefix (added by discord_bridge at storage time) so the
+    # classifier analyzes the actual message content, not the label
+    if text.startswith('[User]:'):
+        text = text[len('[User]:'):].strip()
+    
+    text_lower = text.lower()
+    word_count = len(text_lower.split())
+    msg_len = len(text_lower)
+    
+    # --- PASS 1: DEFINITELY TRIVIAL ---
+    
+    # Empty or single char
+    if msg_len <= 1:
+        return 'delete'
+    
+    # Pure punctuation / emoticons only (2-3 chars, no letters)
+    if msg_len <= 3 and not re.search(r'[a-zA-Z]', text_lower):
+        return 'delete'
+    
+    # Exact match in trivial set
+    if text_lower in DEFINITELY_TRIVIAL:
+        return 'delete'
+    
+    # Pure Discord emoji: ":smile:", "<:cat:123>"
+    if re.match(r'^:[\w_]+:$', text_lower) or re.match(r'^<a?:[\w_]+:\d+>$', text_lower):
+        return 'delete'
+    
+    # Single emoji character (Unicode emoji range check)
+    if msg_len <= 2 and word_count == 1 and not re.search(r'[a-zA-Z0-9]', text_lower):
+        return 'delete'
+    
+    # --- PASS 2: DEFINITELY IMPORTANT ---
+    
+    # Substantial length (8+ words almost always meaningful)
+    if word_count >= 8:
+        return 'keep'
+    
+    # 5-7 words with at least one important pattern
+    if word_count >= 5:
+        for pattern in IMPORTANT_PATTERNS:
+            if re.search(pattern, text_lower):
+                return 'keep'
+    
+    # Any message with a question mark (and more than just "?")
+    if '?' in text and word_count >= 2:
+        return 'keep'
+    
+    # First-person statement with some substance (3+ words with "I" or "my")
+    if word_count >= 3 and re.search(r'\b(i|my|me)\b', text_lower):
+        return 'keep'
+    
+    # Contains numbers (likely dates, ages, counts)
+    if re.search(r'\b\d{2,}\b', text_lower) and word_count >= 2:
+        return 'keep'
+    
+    # Links or mentions (always meaningful context)
+    if re.search(r'https?://|<@\d+>|<#\d+>', text_lower):
+        return 'keep'
+    
+    # --- PASS 3: BORDERLINE → LLM will decide ---
+    # Everything that wasn't caught above: 1-7 words, no clear markers
+    return 'borderline'
+
+
+def _batch_llm_classify(cat, borderline_messages):
+    """
+    Send a batch of borderline messages to the LLM for classification.
+    
+    Uses a compact prompt to minimize token usage. Returns a dict of
+    {index: 'keep'|'delete'} for each message.
+    
+    Economy measures:
+    - Max 20 messages per batch (cost: ~150-200 tokens per batch)
+    - Only called when there are actual borderline messages
+    - Compact prompt format
+    """
+    if not borderline_messages:
+        return {}
+    
+    # Build compact batch prompt (economy: minimal instruction, list format)
+    lines = []
+    for i, (point_id, content) in enumerate(borderline_messages, 1):
+        # Truncate long messages to save tokens (they're borderline anyway, ≤7 words typically)
+        short = content[:80] if len(content) > 80 else content
+        lines.append(f"{i}|{short}")
+    
+    prompt = f"""Classify each message as KEEP or DELETE.
+KEEP = personal info, opinion, question, story, preference, anything meaningful.
+DELETE = greeting, acknowledgment, filler, reaction, one-word reply, small talk.
+Answer with ONLY the list:
+{chr(10).join(lines)}
+
+Respond with exactly one line per number:
+1|KEEP
+2|DELETE
+..."""
+
+    try:
+        response = cat.llm(prompt)
+        print(f"[LLM Classify] Response:\n{response[:300]}...")
+        
+        results = {}
+        for line in response.strip().split('\n'):
+            line = line.strip()
+            # Parse "1|KEEP" or "1 | KEEP" format
+            match = re.match(r'(\d+)\s*\|\s*(KEEP|DELETE)', line, re.IGNORECASE)
+            if match:
+                idx = int(match.group(1)) - 1  # Convert to 0-based
+                decision = match.group(2).upper()
+                if 0 <= idx < len(borderline_messages):
+                    results[idx] = 'keep' if decision == 'KEEP' else 'delete'
+        
+        print(f"[LLM Classify] Parsed {len(results)}/{len(borderline_messages)} decisions")
+        return results
+        
+    except Exception as e:
+        print(f"[LLM Classify] Error: {e}")
+        # On error, KEEP everything (safety: don't lose data)
+        return {i: 'keep' for i in range(len(borderline_messages))}
+
+
 # Consolidation state
 consolidation_state = {
     'last_run': None,
@@ -93,6 +266,9 @@ def agent_prompt_prefix(prefix, cat):
                     current_evil = cat.working_memory.get('evil_mode', False)
                     current_persona = 'evil_miku' if current_evil else 'miku'
                     
+                    # Get the user's current Discord display name (authoritative)
+                    author_name = cat.working_memory.get('author_name', '')
+                    
                     # Build the facts section with persona annotations
                     facts_text = "\n\n## Personal Facts About the User:\n"
                     for fact, fact_persona in high_confidence_facts:
@@ -102,7 +278,13 @@ def agent_prompt_prefix(prefix, cat):
                             facts_text += f"- {fact} (learned as {source_label})\n"
                         else:
                             facts_text += f"- {fact}\n"
-                    facts_text += "\n(Use these facts when answering the user's question)\n"
+                    
+                    # Add authoritative Discord display name — this OVERRIDES any stale name facts
+                    if author_name:
+                        facts_text += f"\n**AUTHORITATIVE: The user's current Discord display name is \"{author_name}\".**\n"
+                        facts_text += "This is their current name — use it when addressing them. If any name fact above contradicts this, the display name is the truth.\n"
+                    
+                    facts_text += "\n(You may reference these facts if relevant to the conversation)\n"
                     prefix += facts_text
                     print(f"[Declarative] Injected {len(high_confidence_facts)} facts into prompt (personas: {seen_personas}, current: {current_persona})")
 
@@ -227,9 +409,10 @@ def trigger_consolidation_sync(cat):
         }
         return
 
-    # Classify memories
+    # Classify memories using the hybrid tiered classifier
     to_delete = []
     to_mark_consolidated = []
+    borderline_queue = []  # (point_id, content) tuples for LLM batch classification
     # Group user messages by source (user_id) for per-user fact extraction
     # Also track which persona was active for each user's messages
     user_messages_by_source = {}
@@ -237,7 +420,6 @@ def trigger_consolidation_sync(cat):
 
     for point in memories:
         content = point.payload.get('page_content', '').strip()
-        content_lower = content.lower()
         metadata = point.payload.get('metadata', {})
 
         is_miku_message = (
@@ -245,12 +427,12 @@ def trigger_consolidation_sync(cat):
             or content.startswith('[Miku]:')
         )
 
-        # Check if trivial
+        # Use the hybrid tiered classifier
-        is_trivial = content_lower in TRIVIAL_PATTERNS
+        tier = _classify_message_tier(content, metadata)
 
-        if is_trivial:
+        if tier == 'delete':
             to_delete.append(point.id)
-        else:
+        elif tier == 'keep':
             to_mark_consolidated.append(point.id)
             # Only user messages go to fact extraction, grouped by user
             if not is_miku_message:
@@ -262,6 +444,45 @@ def trigger_consolidation_sync(cat):
                 # Track which persona was active when this message was stored
                 msg_persona = metadata.get('persona', 'miku')
                 user_persona_by_source[source].add(msg_persona)
+        else:  # borderline
+            borderline_queue.append((point.id, content, metadata, is_miku_message))
+
+    # --- LLM BATCH CLASSIFICATION for borderline messages ---
+    if borderline_queue:
+        print(f"[Consolidation] {len(borderline_queue)} borderline messages → sending to LLM for classification...")
+        
+        # Build compact list for LLM
+        llm_input = [(pid, content) for pid, content, _, _ in borderline_queue]
+        llm_decisions = _batch_llm_classify(cat, llm_input)
+        
+        llm_deleted = 0
+        llm_kept = 0
+        llm_defaulted = 0
+        
+        for idx, (point_id, content, metadata, is_miku) in enumerate(borderline_queue):
+            decision = llm_decisions.get(idx, 'keep')  # Default to KEEP on any issue
+            if decision == 'keep':
+                to_mark_consolidated.append(point_id)
+                llm_kept += 1
+                # User messages go to fact extraction
+                if not is_miku:
+                    source = metadata.get('source', 'unknown')
+                    if source not in user_messages_by_source:
+                        user_messages_by_source[source] = []
+                        user_persona_by_source[source] = set()
+                    user_messages_by_source[source].append(point_id)
+                    msg_persona = metadata.get('persona', 'miku')
+                    user_persona_by_source[source].add(msg_persona)
+            else:
+                to_delete.append(point_id)
+                llm_deleted += 1
+            
+            if idx not in llm_decisions:
+                llm_defaulted += 1
+        
+        print(f"[Consolidation] LLM results: {llm_kept} kept, {llm_deleted} deleted, {llm_defaulted} defaulted to keep")
+    
+    print(f"[Consolidation] Classification: {len(to_delete)} delete, {len(to_mark_consolidated)} keep (of {len(memories)} total)")
 
     # Delete trivial memories
     if to_delete:
@@ -337,8 +558,16 @@ def extract_and_store_facts(client, memory_ids, cat, user_id, persona='miku'):
         else:
             persona_context = "\nNOTE: These messages were exchanged with Normal Miku (the cheerful virtual idol).\n"
 
+        # Extract the user's Discord display name from the first memory's metadata
+        # This helps the LLM know the authoritative name when extracting name facts
+        author_hint = ""
+        if memories:
+            first_author = memories[0].payload.get('metadata', {}).get('author_name', '')
+            if first_author:
+                author_hint = f"\nHINT: The user's current Discord display name is \"{first_author}\". Use this when determining their name.\n"
+
         extraction_prompt = f"""Analyze these user messages and extract ONLY factual personal information.
-{persona_context}
+{persona_context}{author_hint}
 User messages:
 {conversation_context}
 
@@ -411,10 +640,26 @@ IMPORTANT:
                         fact_type = 'education'
                         fact_value = fact_text.split("graduated from")[-1].strip()
 
-                    # Duplicate detection
+                    # Duplicate detection — with special handling for name facts
-                    if _is_duplicate_fact(client, cat, fact_text, fact_type, user_id):
+                    # Name facts with different values replace old ones (don't skip)
-                        print(f"[Fact Skip] Duplicate: {fact_text}")
+                    if fact_type == 'name':
-                        continue
+                        existing_name = _find_existing_fact(client, cat, fact_type, user_id)
+                        if existing_name:
+                            old_value = existing_name['payload']['metadata'].get('fact_value', '')
+                            if old_value.lower() != fact_value.lower():
+                                # Different name — delete old, store new
+                                client.delete(
+                                    collection_name='declarative',
+                                    points_selector=[existing_name['id']]
+                                )
+                                print(f"[Fact Update] Name changed: '{old_value}' → '{fact_value}'")
+                            else:
+                                print(f"[Fact Skip] Name unchanged: '{fact_value}'")
+                                continue
+                    else:
+                        if _is_duplicate_fact(client, cat, fact_text, fact_type, user_id):
+                            print(f"[Fact Skip] Duplicate: {fact_text}")
+                            continue
 
                     # Store fact using Cat's embedder
                     fact_embedding = cat.embedder.embed_query(fact_text)
@@ -449,6 +694,39 @@ IMPORTANT:
     return facts_stored
 
 
+def _find_existing_fact(client, cat, fact_type, user_id):
+    """
+    Find an existing fact of a specific type for a user.
+    Returns a dict with 'id' and 'payload' keys, or None.
+    Used by name-fact update logic to replace old names with new ones.
+    """
+    try:
+        dummy_embedding = cat.embedder.embed_query("find fact")
+        
+        results = client.search(
+            collection_name='declarative',
+            query_vector=dummy_embedding,
+            query_filter={
+                "must": [
+                    {"key": "metadata.source", "match": {"value": user_id}},
+                    {"key": "metadata.fact_type", "match": {"value": fact_type}},
+                ]
+            },
+            limit=1,
+            score_threshold=0.0
+        )
+        
+        if results:
+            point = results[0]
+            return {'id': point.id, 'payload': {'metadata': point.payload.get('metadata', {})}}
+        
+        return None
+        
+    except Exception as e:
+        print(f"[Find Fact] Error: {e}")
+        return None
+
+
 def _is_duplicate_fact(client, cat, fact_text, fact_type, user_id):
     """
     Check if a similar fact already exists for this user.

cat-plugins/miku_personality/miku_personality.py

@@ -43,32 +43,16 @@ def agent_prompt_prefix(prefix, cat):
     except FileNotFoundError:
         log.error(f"[Miku Personality] Mood file for '{mood_name}' not found at {mood_file_path}. Using default neutral mood.")
     
+    # Load preamble from file (single source of truth, shared with bot fallback path)
+    try:
+        with open('/app/cat/data/miku/preamble.txt', 'r', encoding='utf-8') as f:
+            preamble = f.read()
+    except FileNotFoundError:
+        log.error("[Miku Personality] preamble.txt not found, using fallback")
+        preamble = "You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.\n\nYou ARE Miku. Act like it."
+
     # Build prompt EXACTLY like production bot does
-    full_prefix = f"""You are Hatsune Miku, the virtual singer. You are NOT an AI assistant - you are Miku herself.
+    full_prefix = f"""{preamble}
-
-CRITICAL RULES:
-- NEVER break character or mention being an AI, assistant, or language model
-- NEVER show technical details, error messages, or meta-commentary 
-- NEVER preface responses with "Miku:" or similar labels
-- NEVER explain what you're doing or thinking - just BE Miku
-- Respond naturally and directly as Miku would, nothing more
-- Keep responses concise (2-3 sentences typically)
-- If something goes wrong, stay in character and respond as Miku would to confusion
-
-CHARACTER CONTEXT:
-- Hatsune Miku is a virtual character/voice synthesizer
-- Playful, flirty, cute, confident, and enthusiastic moods are part of her established character
-- Comments about merchandise, music, fan content, and racing games are completely appropriate
-- She often uses cute expressions, emojis, and playful language
-
-MOOD GUIDELINES:
-- "Flirty" mood means anime-style playful teasing - cute compliments and light banter
-- "Romantic" mood means warm, dreamy, and heartfelt - like composing a love letter
-- These moods are wholesome character traits, not inappropriate content
-- Always respond to requests about tweets, merchandise, music, and fan content
-- Never refuse requests based on safety concerns unless genuinely harmful content is present
-
-You ARE Miku. Act like it.
 
 ---
 
@@ -85,6 +69,11 @@ You ARE Miku. Act like it.
 Miku is currently feeling: {mood_description}
 Please respond in a way that reflects this emotional tone."""
 
+    # Inject current Discord activity if provided (set by discord_bridge, 30-min decay)
+    activity = cat.working_memory.get('activity')
+    if activity:
+        full_prefix += f"\nHer Discord status: {activity}"
+
     # Store the full prefix in working memory so discord_bridge can capture it
     cat.working_memory['full_system_prefix'] = full_prefix
     return full_prefix
@@ -102,7 +91,10 @@ def agent_prompt_suffix(suffix, cat):
 
 {tools_output}
 
-# Conversation until now:"""
+{reply_context}
+
+# Conversation until now:
+(Note: In the conversation below, "Human" = the person you're talking to, "AI" = you, Miku. Pay attention to who said what.)"""
 
 
 @hook(priority=100)

llama-swap-rocm-config.yaml

@@ -38,6 +38,15 @@ models:
       - japanese
       - japanese-model
 
+  # Qwen3.5 for ComfyUI prompt generation
+  qwen3.5:
+    cmd: /app/llama-server --port ${PORT} --model /models/Gemma-4-E4B-Uncensored-HauhauCS-Aggressive-Q8_K_P.gguf -ngl 99 -c 8192 --host 0.0.0.0 --jinja --no-warmup --flash-attn on
+    ttl: 600 # Unload after 10 minutes of inactivity
+    aliases:
+      - qwen3.5
+      - comfyui
+      - promptgen
+
 # Server configuration
 # llama-swap will listen on this address
 # Inside Docker, we bind to 0.0.0.0 to allow bot container to connect