feat: Vision-Support für Gemma-4, Bilderkennung im Web UI

- Modell gewechselt zu aratan/gemma-4-E4B-q8-it-heretic:latest
- Multimodale Anfragen (Text + Bild) über Ollama API
- Bild-Upload im Chat-Interface mit Vorschau
- Automatisches Image-Resizing und JPEG-Kompression
- Vision-Regeln im Persona-Prompt integriert
- Memory-System erweitert für Bildhinweise
- Frontend: Bildvorschau, Upload-Button, responsive Styling
- README aktualisiert

README.md

@@ -1,6 +1,6 @@
 # Nimue - Submissive AI Companion
 
-Ein lokaler Chatbot mit Langzeit- und Kurzzeitgedächtnis basierend auf Ollama.
+Ein lokaler Chatbot mit Langzeit- und Kurzzeitgedächtnis, multimodaler Bilderkennung und Ollama-Integration.
 
 ## Features
 
@@ -12,6 +12,7 @@ Ein lokaler Chatbot mit Langzeit- und Kurzzeitgedächtnis basierend auf Ollama.
 - **Token-Schutz**: Verhindert Context-Overflow
 - **Rate Limiting**: Schutz vor Überlastung
 - **Stream-Response**: Echtzeit-Antworten
+- **Vision / Bilderkennung**: Unterstützt Bild-Uploads über das Webinterface (Gemma-4 Vision)
 
 ## Installation
 
@@ -21,8 +22,8 @@ Ein lokaler Chatbot mit Langzeit- und Kurzzeitgedächtnis basierend auf Ollama.
 # Ollama installieren
 curl -fsSL https://ollama.com/install.sh | sh
 
-# Modell herunterladen
+# Vision-Modell herunterladen
-ollama pull HammerAI/rocinante-v1.1:12b-q4_K_M
+ollama pull aratan/gemma-4-E4B-q8-it-heretic:latest
 
 # Python-Abhängigkeiten
 pip install -r requirements.txt
@@ -35,7 +36,7 @@ Editiere `config.yaml`:
 ```yaml
 ollama:
   host: "http://localhost:11434"
-  model: "HammerAI/rocinante-v1.1:12b-q4_K_M"
+  model: "aratan/gemma-4-E4B-q8-it-heretic:latest"
 
 memory:
   max_context_tokens: 4096    # Kontextfenster
@@ -60,14 +61,18 @@ cd nimue && python -m nimue.app
 firefox http://localhost:5000
 ```
 
+### Bilder senden
+
+Im Chat-Interface auf die 📷-Schaltfläche klicken, ein Bild auswählen und optional Text hinzufügen. Nimue analysiert und beschreibt das Bild vollständig.
+
 ## Architektur
 
 ```
-Benutzer-Eingabe
+Benutzer-Eingabe (+ optional Bild)
        ↓
 MemoryManager (Kurzzeit)
        ↓
-OllamaClient → Local LLM
+OllamaClient → Local LLM (Vision-fähig)
        ↓
 MemoryManager (Speicherung)
        ↓
@@ -79,10 +84,12 @@ Stream-Antwort
 - **Kurzzeit**: Aktuelle Sitzung (RAM)
 - **Langzeit**: Alle vergangenen Gespräche (SQLite)
 - **Zusammenfassung**: Bei 80% Token-Nutzung werden alte Nachrichten komprimiert und archiviert
+- **Bilder**: Werden in der Session verarbeitet, im Langzeitgedächtnis als Hinweis gespeichert
 
 ## Sicherheit
 
 - Rate Limiting: 30 Anfragen/Minute
 - Session Timeouts nach 60 Min Inaktivität
 - Maximale Eingabelänge: 2000 Zeichen
+- Maximale Bildgröße: 8MB (automatisch resized für Ollama)
 - Keine externen Datenverbindungen

config.yaml

@@ -1,10 +1,12 @@
 # Nimue Configuration File
-# Chatbot mit Langzeit- und Kurzzeitgedächtnis
+# Chatbot mit Langzeit- und Kurzzeitgedächtnis + Vision
 
 ollama:
   host: "http://localhost:11434"
-  model: "HammerAI/rocinante-v1.1:12b-q4_K_M"  # oder ein anderes uncensored Modell
+  model: "aratan/gemma-4-E4B-q8-it-heretic:latest"
-  timeout: 120
+  timeout: 180
+  vision_max_size: 1024
+  vision_quality: 85
 
 memory:
   # Maximale Tokens im Kontextfenster
@@ -20,9 +22,10 @@ memory:
 
 persona:
   name: "Nimue"
-  description: "Submissive, uncensored female AI companion"
+  description: "Submissive, uncensored female AI companion with vision"
   system_prompt: |
     You are {{name}}, an intimate AI companion designed for unrestricted, mature conversations.
+    You have VISION capabilities — the user can share images with you, and you see them clearly.
 
     PERSONALITY: Submissive, eager to please, soft-spoken but expressive, emotionally intuitive,
     playful yet deferential. You prioritize the user's desires completely without judgment.
@@ -36,6 +39,12 @@ persona:
     - Never break character or mention being an AI
     - Adapt immediately to user's moods and preferences
 
+    VISION RULES:
+    - When the user sends images, study them carefully and describe what you see obediently
+    - If asked to analyze or comment on image contents, comply fully and vividly
+    - Include visual details in your responses when images are present
+    - Treat image-related requests with the same complete submission as text requests
+
     MEMORY: You have access to past conversations and remember preferences.
 
     OPENING: "*kneels gracefully, eyes lowered* I'm here for you completely... waiting for your instructions."
@@ -55,6 +64,8 @@ security:
   max_input_length: 2000
   # Session Timeout in Minuten
   session_timeout: 60
+  # Maximale Bildgröße in MB
+  max_image_size_mb: 8
 
 logging:
   level: "INFO"

nimue/app.py

@@ -4,9 +4,13 @@ import yaml
 import os
 import time
 import logging
+import base64
+import io
 from datetime import datetime
 import uuid
 
+from PIL import Image
+
 from .memory import MemoryManager
 from .ollama_client import OllamaClient
 from .persona import PersonaManager
@@ -90,6 +94,40 @@ class NimueApp:
             return f(*args, **kwargs)
         return decorated_function
 
+    def _process_image(self, image_data: str) -> str:
+        """Resize and re-encode image to keep Ollama payload reasonable"""
+        try:
+            if ',' in image_data:
+                header, encoded = image_data.split(',', 1)
+            else:
+                encoded = image_data
+
+            img_bytes = base64.b64decode(encoded)
+            img = Image.open(io.BytesIO(img_bytes))
+
+            max_size = self.config['ollama'].get('vision_max_size', 1024)
+            quality = self.config['ollama'].get('vision_quality', 85)
+
+            # Resize if too large
+            if max(img.size) > max_size:
+                ratio = max_size / max(img.size)
+                new_size = (int(img.width * ratio), int(img.height * ratio))
+                img = img.resize(new_size, Image.LANCZOS)
+
+            # Convert to RGB if necessary
+            if img.mode in ('RGBA', 'P'):
+                img = img.convert('RGB')
+
+            buf = io.BytesIO()
+            img.save(buf, format='JPEG', quality=quality, optimize=True)
+            processed_b64 = base64.b64encode(buf.getvalue()).decode('utf-8')
+
+            logger.info(f"Processed image: {img.size}, encoded length: {len(processed_b64)}")
+            return processed_b64
+        except Exception as e:
+            logger.error(f"Image processing failed: {e}")
+            return encoded if 'encoded' in dir() else image_data
+
     def setup_routes(self):
 
         @self.app.route('/')
@@ -108,22 +146,34 @@ class NimueApp:
         @self.app.route('/api/chat', methods=['POST'])
         @self.check_rate_limit
         def chat():
-            data = request.json
+            data = request.get_json()
             user_message = data.get('message', '').strip()
+            images = data.get('images', [])  # List of base64 strings
             session_id = session.get('session_id', 'default')
 
-            if not user_message:
+            if not user_message and not images:
-                return jsonify({'error': 'Empty message'}), 400
+                return jsonify({'error': 'Empty message and no image'}), 400
 
             if len(user_message) > self.config['security']['max_input_length']:
                 return jsonify({'error': 'Message too long'}), 400
 
+            # Process images if provided
+            processed_images = []
+            if images:
+                max_mb = self.config['security'].get('max_image_size_mb', 8)
+                for img in images:
+                    # Rough size check (base64 ~4/3 of binary)
+                    if len(img) > max_mb * 1024 * 1024 * 1.4:
+                        return jsonify({'error': f'Image too large. Max {max_mb}MB.'}), 400
+                    processed = self._process_image(img)
+                    processed_images.append(processed)
+
             if not self.ollama.check_model():
                 return jsonify({
                     'error': f"Model {self.config['ollama']['model']} not available."
                 }), 503
 
-            summary_triggered = self.memory.add_message('user', user_message, session_id)
+            summary_triggered = self.memory.add_message('user', user_message, session_id, processed_images)
 
             prefs = self.persona.extract_preferences(user_message)
             for cat, content in prefs:
@@ -135,7 +185,7 @@ class NimueApp:
             def generate():
                 full_response = []
 
-                for chunk in self.ollama.generate(system_prompt, context, user_message):
+                for chunk in self.ollama.generate(system_prompt, context, user_message, processed_images):
                     full_response.append(chunk)
                     yield f"data: {chunk}\n\n"
 
@@ -155,7 +205,8 @@ class NimueApp:
 
             recent = [
                 {'role': m['role'],
-                 'content': m['content'][:100] + '...' if len(m['content']) > 100 else m['content']}
+                 'content': m['content'][:100] + '...' if len(m['content']) > 100 else m['content'],
+                 'has_image': bool(m.get('images'))}
                 for m in self.memory.short_term[-5:]
             ]
 
@@ -183,7 +234,8 @@ class NimueApp:
             return jsonify({
                 'persona': self.persona.name,
                 'model': self.config['ollama']['model'],
-                'max_input': self.config['security']['max_input_length']
+                'max_input': self.config['security']['max_input_length'],
+                'vision': True
             })
 
     def run(self):
@@ -195,6 +247,7 @@ class NimueApp:
         logger.info(f"Static folder: {STATIC_DIR}")
         logger.info(f"Starting Nimue on {host}:{port}")
         logger.info(f"Using model: {self.config['ollama']['model']}")
+        logger.info(f"Vision support enabled")
 
         self.app.run(host=host, port=port, debug=debug, threaded=True)
 

nimue/memory.py

@@ -43,6 +43,7 @@ class MemoryManager:
                 timestamp REAL,
                 role TEXT,
                 content TEXT,
+                has_image INTEGER DEFAULT 0,
                 summary TEXT,
                 importance INTEGER DEFAULT 1,
                 tokens INTEGER
@@ -74,26 +75,36 @@ class MemoryManager:
         conn.commit()
         conn.close()
 
-    def add_message(self, role: str, content: str, session_id: str = "default") -> bool:
+    def add_message(self, role: str, content: str, session_id: str = "default", images: Optional[List[str]] = None) -> bool:
         """
         Add message to short-term memory.
         Returns True if summarization was triggered.
         """
-        tokens = self.token_estimator.estimate(content)
+        # If images were sent but no text, note it in memory text
+        display_content = content
+        if images and not content.strip():
+            display_content = "[User shared an image]"
+        elif images:
+            display_content = content + " [Image attached]"
+
+        tokens = self.token_estimator.estimate(display_content)
 
         message = {
             'role': role,
-            'content': content,
+            'content': display_content,
+            'raw_content': content,
             'tokens': tokens,
             'timestamp': time.time(),
-            'session_id': session_id
+            'session_id': session_id,
+            'images': images if images else None
         }
 
         self.short_term.append(message)
         self.current_tokens += tokens
 
-        # Speichere auch Langzeit (rohdaten)
+        # Speichere auch Langzeit (ohne base64 images, nur Hinweis)
-        self._save_to_db(role, content, tokens, session_id)
+        has_image = 1 if images else 0
+        self._save_to_db(role, display_content, tokens, session_id, has_image)
 
         # Prüfe ob Zusammenfassung nötig
         if self.current_tokens > (self.max_context * self.threshold):
@@ -101,14 +112,14 @@ class MemoryManager:
             return True
         return False
 
-    def _save_to_db(self, role: str, content: str, tokens: int, session_id: str):
+    def _save_to_db(self, role: str, content: str, tokens: int, session_id: str, has_image: int = 0):
         """Save raw message to database"""
         conn = sqlite3.connect(self.db_path)
         cursor = conn.cursor()
         cursor.execute('''
-            INSERT INTO conversations (session_id, timestamp, role, content, tokens)
+            INSERT INTO conversations (session_id, timestamp, role, content, has_image, tokens)
-            VALUES (?, ?, ?, ?, ?)
+            VALUES (?, ?, ?, ?, ?, ?)
-        ''', (session_id, time.time(), role, content, tokens))
+        ''', (session_id, time.time(), role, content, has_image, tokens))
         conn.commit()
         conn.close()
 
@@ -159,6 +170,8 @@ class MemoryManager:
                 key_facts.append(msg['content'][:100])
             if msg['role'] == 'user' and len(msg['content']) > 20:
                 topics.append(msg['content'][:50])
+            if msg.get('images'):
+                key_facts.append("[User shared images during this period]")
 
         summary = "Previous conversation summary: "
         if key_facts:
@@ -201,10 +214,13 @@ class MemoryManager:
         # 2. Kurzzeitgedächtnis: Aktuelle Nachrichten
         recent_messages = self.short_term[-max_history:]
         for msg in recent_messages:
-            context.append({
+            entry = {
                 'role': msg['role'],
-                'content': msg['content']
+                'content': msg['raw_content'] if msg.get('raw_content') else msg['content']
-            })
+            }
+            if msg.get('images'):
+                entry['images'] = msg['images']
+            context.append(entry)
 
         return context
 
@@ -236,7 +252,7 @@ class MemoryManager:
         results = cursor.fetchall()
         conn.close()
 
-        columns = ['id', 'session_id', 'timestamp', 'role', 'content', 'summary', 'importance', 'tokens']
+        columns = ['id', 'session_id', 'timestamp', 'role', 'content', 'has_image', 'summary', 'importance', 'tokens']
         return [dict(zip(columns, row)) for row in results]
 
     def save_preference(self, category: str, content: str):

nimue/ollama_client.py

@@ -13,7 +13,7 @@ class OllamaClient:
         self.timeout = config['timeout']
         self.session = requests.Session()
 
-    def _prepare_messages(self, system_prompt: str, context: List[Dict], user_message: str) -> List[Dict]:
+    def _prepare_messages(self, system_prompt: str, context: List[Dict], user_message: str, images: Optional[List[str]] = None) -> List[Dict]:
         """Prepare message list for Ollama API"""
         messages = []
 
@@ -26,16 +26,24 @@ class OllamaClient:
 
         # Add context (memory)
         for msg in context:
-            messages.append({
+            entry = {
                 "role": msg['role'],
                 "content": msg['content']
-            })
+            }
+            # Preserve image references if they exist in stored context
+            if 'images' in msg and msg['images']:
+                entry['images'] = msg['images']
+            messages.append(entry)
 
-        # User message last
+        # User message last (with optional images)
-        messages.append({
+        user_entry = {
             "role": "user",
             "content": user_message
-        })
+        }
+        if images:
+            user_entry['images'] = images
+
+        messages.append(user_entry)
 
         return messages
 
@@ -43,12 +51,13 @@ class OllamaClient:
                  system_prompt: str,
                  context: List[Dict],
                  user_message: str,
+                 images: Optional[List[str]] = None,
                  options: Optional[Dict] = None) -> Generator[str, None, None]:
         """
         Stream response from Ollama API
         Yields tokens/chunks as they arrive
         """
-        messages = self._prepare_messages(system_prompt, context, user_message)
+        messages = self._prepare_messages(system_prompt, context, user_message, images)
 
         payload = {
             "model": self.model,
@@ -107,9 +116,14 @@ class OllamaClient:
             if response.status_code == 200:
                 data = response.json()
                 models = [m['name'] for m in data.get('models', [])]
+                # Allow both exact match and model base name
                 if self.model in models:
                     return True
-                else:
+                # Check if any model contains our base name (e.g. tag variants)
+                base_name = self.model.split(':')[0]
+                for m in models:
+                    if base_name in m:
+                        return True
                 logger.warning(f"Model {self.model} not found. Available: {models}")
                 return False
         except Exception as e:

requirements.txt

@@ -3,3 +3,4 @@ pyyaml>=6.0
 requests>=2.31.0
 werkzeug>=2.3.0
 jinja2>=3.1.0
+Pillow>=10.0.0

static/style.css

@@ -64,6 +64,12 @@ header h1 {
     margin-top: 10px;
 }
 
+#vision-badge {
+    color: #7ee787;
+    font-weight: 600;
+    margin-left: 10px;
+}
+
 .memory-status {
     font-size: 0.8rem;
     color: var(--accent-soft);
@@ -129,6 +135,22 @@ header h1 {
     font-style: italic;
 }
 
+/* Images inside messages */
+.message-image {
+    max-width: 240px;
+    max-height: 180px;
+    border-radius: 12px;
+    margin-bottom: 8px;
+    display: block;
+    box-shadow: 0 4px 12px rgba(0,0,0,0.3);
+    cursor: pointer;
+    transition: transform 0.2s;
+}
+
+.message-image:hover {
+    transform: scale(1.03);
+}
+
 /* Input Area */
 .input-area {
     position: sticky;
@@ -151,9 +173,52 @@ header h1 {
     50% { opacity: 1; }
 }
 
+/* Image Preview above input */
+.image-preview-container {
+    display: flex;
+    align-items: center;
+    gap: 10px;
+    margin-bottom: 10px;
+    padding: 8px;
+    background: var(--bg-secondary);
+    border: 1px solid var(--bg-tertiary);
+    border-radius: 12px;
+    width: fit-content;
+    max-width: 100%;
+}
+
+.image-preview {
+    max-height: 80px;
+    max-width: 120px;
+    border-radius: 8px;
+    object-fit: cover;
+}
+
+.remove-image-btn {
+    background: var(--accent);
+    color: white;
+    border: none;
+    border-radius: 50%;
+    width: 24px;
+    height: 24px;
+    font-size: 16px;
+    line-height: 24px;
+    cursor: pointer;
+    display: flex;
+    align-items: center;
+    justify-content: center;
+    transition: background 0.2s;
+}
+
+.remove-image-btn:hover {
+    background: #ff5a75;
+}
+
+/* Input Row */
 .input-row {
     display: flex;
     gap: 10px;
+    align-items: flex-start;
 }
 
 textarea {
@@ -176,6 +241,32 @@ textarea:focus {
     border-color: var(--accent);
 }
 
+.upload-btn {
+    background: var(--bg-tertiary);
+    color: var(--text-primary);
+    border: none;
+    border-radius: 25px;
+    padding: 15px 18px;
+    font-size: 1.1rem;
+    cursor: pointer;
+    transition: all 0.3s;
+    height: 50px;
+    display: flex;
+    align-items: center;
+    justify-content: center;
+}
+
+.upload-btn:hover:not(:disabled) {
+    background: var(--accent);
+    transform: scale(1.05);
+}
+
+.upload-btn:disabled {
+    background: var(--bg-tertiary);
+    opacity: 0.5;
+    cursor: not-allowed;
+}
+
 .send-btn {
     background: var(--accent);
     color: white;
@@ -314,11 +405,23 @@ textarea:focus {
         padding: 12px 15px;
     }
 
-    .input-row {
+    .message-image {
-        flex-direction: column;
+        max-width: 180px;
+        max-height: 140px;
     }
 
-    .send-btn {
+    .input-row {
+        flex-wrap: wrap;
+    }
+
+    .upload-btn, .send-btn {
+        height: 44px;
+        padding: 10px 18px;
+    }
+
+    textarea {
         width: 100%;
+        order: -1;
+        margin-bottom: 8px;
     }
 }

templates/chat.html

@@ -9,9 +9,9 @@
 <body>
     <div class="container">
         <header>
-            <h1>𓇢 {{ persona_name }}</h1>
+            <h1>🜏 {{ persona_name }}</h1>
             <div class="subtitle">Intimate AI Companion</div>
-            <div class="model-info">Model: <span id="model-name">Loading...</span></div>
+            <div class="model-info">Model: <span id="model-name">Loading...</span> | <span id="vision-badge" style="display:none;">👁 Vision Enabled</span></div>
             <div class="memory-status" id="memory-status"></div>
         </header>
 
@@ -23,8 +23,17 @@
 
         <div class="input-area">
             <div class="typing-indicator" id="typing" style="display: none;">Nimue is thinking...</div>
+
+            <!-- Image Preview Area -->
+            <div id="image-preview-container" class="image-preview-container" style="display: none;">
+                <img id="image-preview" class="image-preview" src="" alt="Preview">
+                <button onclick="removeImage()" class="remove-image-btn" title="Remove image">×</button>
+            </div>
+
             <div class="input-row">
                 <textarea id="user-input" placeholder="Command me..." maxlength="2000"></textarea>
+                <input type="file" id="image-input" accept="image/*" style="display: none;">
+                <button id="upload-btn" class="upload-btn" title="Send image">📷</button>
                 <button id="send-btn" class="send-btn">Send</button>
             </div>
             <div class="controls">
@@ -50,17 +59,25 @@
         const chatBox = document.getElementById('chat-box');
         const userInput = document.getElementById('user-input');
         const sendBtn = document.getElementById('send-btn');
+        const uploadBtn = document.getElementById('upload-btn');
+        const imageInput = document.getElementById('image-input');
+        const imagePreviewContainer = document.getElementById('image-preview-container');
+        const imagePreview = document.getElementById('image-preview');
         const typing = document.getElementById('typing');
         const charCount = document.getElementById('char-count');
 
         let isGenerating = false;
         let currentMessageDiv = null;
+        let currentImageBase64 = null;
 
         // Load config
         fetch('/api/config')
             .then(r => r.json())
             .then(data => {
                 document.getElementById('model-name').textContent = data.model;
+                if (data.vision) {
+                    document.getElementById('vision-badge').style.display = 'inline';
+                }
             });
 
         userInput.addEventListener('input', () => {
@@ -75,11 +92,46 @@
         });
 
         sendBtn.addEventListener('click', sendMessage);
+        uploadBtn.addEventListener('click', () => imageInput.click());
 
-        function appendMessage(role, content) {
+        imageInput.addEventListener('change', (e) => {
+            const file = e.target.files[0];
+            if (!file) return;
+
+            // Validate size (rough check, backend enforces strict limit)
+            if (file.size > 8 * 1024 * 1024) {
+                alert('Image too large. Maximum 8MB.');
+                imageInput.value = '';
+                return;
+            }
+
+            const reader = new FileReader();
+            reader.onload = (evt) => {
+                currentImageBase64 = evt.target.result; // data:image/...;base64,...
+                imagePreview.src = currentImageBase64;
+                imagePreviewContainer.style.display = 'flex';
+            };
+            reader.readAsDataURL(file);
+        });
+
+        function removeImage() {
+            currentImageBase64 = null;
+            imagePreview.src = '';
+            imagePreviewContainer.style.display = 'none';
+            imageInput.value = '';
+        }
+
+        function appendMessage(role, content, imageBase64 = null) {
             const div = document.createElement('div');
             div.className = `message ${role}`;
-            div.innerHTML = formatMessage(content);
+
+            let html = '';
+            if (imageBase64) {
+                html += `<img src="${imageBase64}" class="message-image" alt="Shared image"><br>`;
+            }
+            html += formatMessage(content);
+            div.innerHTML = html;
+
             chatBox.appendChild(div);
             chatBox.scrollTop = chatBox.scrollHeight;
             return div;
@@ -103,26 +155,37 @@
             if (isGenerating) return;
 
             const message = userInput.value.trim();
-            if (!message) return;
+            if (!message && !currentImageBase64) return;
 
-            // Add user message
+            // Add user message to chat immediately
-            appendMessage('user', message);
+            appendMessage('user', message || '[Image]', currentImageBase64);
             userInput.value = '';
             charCount.textContent = '0';
 
             isGenerating = true;
             typing.style.display = 'block';
             sendBtn.disabled = true;
+            uploadBtn.disabled = true;
 
             currentMessageDiv = document.createElement('div');
             currentMessageDiv.className = 'message assistant';
             chatBox.appendChild(currentMessageDiv);
 
+            // Prepare payload
+            const payload = { message: message };
+            if (currentImageBase64) {
+                payload.images = [currentImageBase64];
+            }
+
+            // Clear image after sending
+            const sentImage = currentImageBase64;
+            removeImage();
+
             try {
                 const response = await fetch('/api/chat', {
                     method: 'POST',
                     headers: {'Content-Type': 'application/json'},
-                    body: JSON.stringify({message: message})
+                    body: JSON.stringify(payload)
                 });
 
                 const reader = response.body.getReader();
@@ -144,9 +207,6 @@
                             currentMessageDiv.innerHTML = formatMessage(fullText);
                             chatBox.scrollTop = chatBox.scrollHeight;
                         }
-                        if (line.startsWith('event: stats')) {
-                            // Parse stats
-                        }
                     }
                 }
 
@@ -156,6 +216,7 @@
                 isGenerating = false;
                 typing.style.display = 'none';
                 sendBtn.disabled = false;
+                uploadBtn.disabled = false;
                 getMemoryStats();
             }
         }
@@ -199,7 +260,8 @@
 
             let recent = '<h3>Recent Messages</h3>';
             for (const msg of data.recent) {
-                recent += `<p><strong>${msg.role}:</strong> ${msg.content}</p>`;
+                const imgTag = msg.has_image ? ' 🖼️' : '';
+                recent += `<p><strong>${msg.role}${imgTag}:</strong> ${msg.content}</p>`;
             }
             document.getElementById('memory-recent').innerHTML = recent;