Checkpoint, I have tools working, but there is a better way to do this.

2024-09-28 11:12:43 -04:00
15 changed files with 342 additions and 1823 deletions
--- a/.gitea/workflows/datadog-sca.yml
+++ b/.gitea/workflows/datadog-sca.yml
@ -1,20 +0,0 @@
-on: [push]
-
-name: Datadog Software Composition Analysis
-
-jobs:
-  software-composition-analysis:
-    runs-on: ubuntu-latest
-    name: Datadog SBOM Generation and Upload
-    steps:
-    - name: Checkout
-      uses: actions/checkout@v3
-    - name: Check imported libraries are secure and compliant
-      id: datadog-software-composition-analysis
-      uses: DataDog/datadog-sca-github-action@main
-      with:
-        dd_api_key: ${{ secrets.DD_API_KEY }}
-        dd_app_key: ${{ secrets.DD_APP_KEY }}
-        dd_service: jarvis
-        dd_env: ci
-        dd_site: us5.datadoghq.com
--- a/.gitea/workflows/datadog-static-analysis.yml
+++ b/.gitea/workflows/datadog-static-analysis.yml
@ -1,21 +0,0 @@
-on: [push]
-
-name: Datadog Static Analysis
-
-jobs:
-  static-analysis:
-    runs-on: ubuntu-latest
-    name: Datadog Static Analyzer
-    steps:
-    - name: Checkout
-      uses: actions/checkout@v3
-    - name: Check code meets quality and security standards
-      id: datadog-static-analysis
-      uses: DataDog/datadog-static-analyzer-github-action@v1
-      with:
-        dd_api_key: ${{ secrets.DD_API_KEY }}
-        dd_app_key: ${{ secrets.DD_APP_KEY }}
-        dd_service: jarvis
-        dd_env: ci
-        dd_site: us5.datadoghq.com
-        cpu_count: 2
--- a/.github/workflows/datadog-sca.yml
+++ b/.github/workflows/datadog-sca.yml
@ -1,20 +0,0 @@
-on: [push]
-
-name: Datadog Software Composition Analysis
-
-jobs:
-  software-composition-analysis:
-    runs-on: ubuntu-latest
-    name: Datadog SBOM Generation and Upload
-    steps:
-    - name: Checkout
-      uses: actions/checkout@v3
-    - name: Check imported libraries are secure and compliant
-      id: datadog-software-composition-analysis
-      uses: DataDog/datadog-sca-github-action@main
-      with:
-        dd_api_key: ${{ secrets.DD_API_KEY }}
-        dd_app_key: ${{ secrets.DD_APP_KEY }}
-        dd_service: jarvis
-        dd_env: ci
-        dd_site: us5.datadoghq.com
--- a/.github/workflows/datadog-static-analysis.yml
+++ b/.github/workflows/datadog-static-analysis.yml
@ -1,21 +0,0 @@
-on: [push]
-
-name: Datadog Static Analysis
-
-jobs:
-  static-analysis:
-    runs-on: ubuntu-latest
-    name: Datadog Static Analyzer
-    steps:
-    - name: Checkout
-      uses: actions/checkout@v3
-    - name: Check code meets quality and security standards
-      id: datadog-static-analysis
-      uses: DataDog/datadog-static-analyzer-github-action@v1
-      with:
-        dd_api_key: ${{ secrets.DD_API_KEY }}
-        dd_app_key: ${{ secrets.DD_APP_KEY }}
-        dd_service: jarvis
-        dd_env: ci
-        dd_site: us5.datadoghq.com
-        cpu_count: 2
--- a/.gitignore
+++ b/.gitignore
@ -174,46 +174,3 @@ cython_debug/
 pyvenv.cfg
 .venv
 pip-selfcheck.json
-
-
-# Logs
-logs
-*.log
-npm-debug.log*
-
-# Runtime data
-pids
-*.pid
-*.seed
-
-# Directory for instrumented libs generated by jscoverage/JSCover
-lib-cov
-
-# Coverage directory used by tools like istanbul
-coverage
-
-# nyc test coverage
-.nyc_output
-
-# Grunt intermediate storage (http://gruntjs.com/creating-plugins#storing-task-files)
-.grunt
-
-# node-waf configuration
-.lock-wscript
-
-# Compiled binary addons (http://nodejs.org/api/addons.html)
-build/Release
-
-# Dependency directories
-node_modules
-jspm_packages
-
-# Optional npm cache directory
-.npm
-
-# Optional REPL history
-.node_repl_history
-.next
-
-config.ini
-*.db
--- a/20
+++ b/20
@ -1,20 +0,0 @@
-# Use an official Python runtime as a parent image
-FROM python:3.9-slim
-
-# Set the working directory in the container
-WORKDIR /app
-
-# Copy the current directory contents into the container at /app
-COPY . /app
-
-# Install any needed packages specified in requirements.txt
-RUN pip install --no-cache-dir -r requirements.txt
-
-# Make port 5001 available to the world outside this container
-EXPOSE 5001
-
-# Define environment variable
-ENV FLASK_APP=main.py
-
-# Run app.py when the container launches
-CMD ["python", "main.py"]
--- a/README.md
+++ b/README.md
@ -1,5 +0,0 @@
-# Jarvis
-
-it's actually not that smart!
-
-
--- a/client.py
+++ b/client.py
@ -1,138 +0,0 @@
-import time
-
-import requests
-
-
-class LLMChatClient:
-    def __init__(self, base_url, api_key):
-        self.base_url = base_url.rstrip("/")
-        self.api_key = api_key
-        self.headers = {"X-API-Key": api_key, "Content-Type": "application/json"}
-
-    def submit_query(self, message):
-        """
-        Submit a query to the LLM Chat Server.
-
-        Args:
-            message (str): The message to send to the server.
-
-        Returns:
-            str: The query ID for the submitted query.
-
-        Raises:
-            requests.RequestException: If the request fails.
-
-        Example:
-            client = LLMChatClient('http://localhost:5001', 'your-api-key')
-            query_id = client.submit_query('What is the capital of France?')
-            print(f"Query ID: {query_id}")
-
-        cURL equivalent:
-            curl -X POST http://localhost:5001/api/v1/query \
-                 -H "Content-Type: application/json" \
-                 -H "X-API-Key: your-api-key" \
-                 -d '{"message": "What is the capital of France?"}'
-        """
-        url = f"{self.base_url}/api/v1/query"
-        data = {"message": message}
-        response = requests.post(url, json=data, headers=self.headers)
-        response.raise_for_status()
-        return response.json()["query_id"]
-
-    def get_query_status(self, query_id):
-        """
-        Get the status of a submitted query.
-
-        Args:
-            query_id (str): The ID of the query to check.
-
-        Returns:
-            dict: A dictionary containing the status and conversation history (if completed).
-
-        Raises:
-            requests.RequestException: If the request fails.
-
-        Example:
-            client = LLMChatClient('http://localhost:5001', 'your-api-key')
-            status = client.get_query_status('query-id-here')
-            print(f"Query status: {status['status']}")
-            if status['status'] == 'completed':
-                print(f"Conversation history: {status['conversation_history']}")
-
-        cURL equivalent:
-            curl -X GET http://localhost:5001/api/v1/query_status/query-id-here \
-                 -H "X-API-Key: your-api-key"
-        """
-        url = f"{self.base_url}/api/v1/query_status/{query_id}"
-        response = requests.get(url, headers=self.headers)
-        response.raise_for_status()
-        return response.json()
-
-    def submit_query_and_wait(self, message, max_wait_time=300, poll_interval=2):
-        """
-        Submit a query and wait for the result.
-
-        Args:
-            message (str): The message to send to the server.
-            max_wait_time (int): Maximum time to wait for the result in seconds.
-            poll_interval (int): Time between status checks in seconds.
-
-        Returns:
-            dict: The completed conversation history.
-
-        Raises:
-            requests.RequestException: If the request fails.
-            TimeoutError: If the query doesn't complete within max_wait_time.
-
-        Example:
-            client = LLMChatClient('http://localhost:5001', 'your-api-key')
-            result = client.submit_query_and_wait('What is the capital of France?')
-            print(f"Conversation history: {result}")
-        """
-        query_id = self.submit_query(message)
-        start_time = time.time()
-
-        while time.time() - start_time < max_wait_time:
-            status = self.get_query_status(query_id)
-            if status["status"] == "completed":
-                return status["conversation_history"]
-            time.sleep(poll_interval)
-
-        raise TimeoutError(f"Query did not complete within {max_wait_time} seconds")
-
-
-class LLMChatAdminClient:
-    def __init__(self, base_url, admin_key):
-        self.base_url = base_url.rstrip("/")
-        self.admin_key = admin_key
-        self.headers = {"X-Admin-Key": admin_key, "Content-Type": "application/json"}
-
-    def generate_api_key(self, username):
-        """
-        Generate a new API key for a user.
-
-        Args:
-            username (str): The username to generate the API key for.
-
-        Returns:
-            dict: A dictionary containing the username and generated API key.
-
-        Raises:
-            requests.RequestException: If the request fails.
-
-        Example:
-            admin_client = LLMChatAdminClient('http://localhost:5001', 'your-admin-key')
-            result = admin_client.generate_api_key('new_user')
-            print(f"Generated API key for {result['username']}: {result['api_key']}")
-
-        cURL equivalent:
-            curl -X POST http://localhost:5001/admin/generate_key \
-                 -H "Content-Type: application/json" \
-                 -H "X-Admin-Key: your-admin-key" \
-                 -d '{"username": "new_user"}'
-        """
-        url = f"{self.base_url}/admin/generate_key"
-        data = {"username": username}
-        response = requests.post(url, json=data, headers=self.headers)
-        response.raise_for_status()
-        return response.json()
--- a/docker-compose.yml
+++ b/docker-compose.yml
@ -1,16 +0,0 @@
-version: '3.8'
-
-services:
-  llm-chat-server:
-    build: .
-    ports:
-      - "5001:5001"
-    volumes:
-      - ./llm_chat_server.db:/app/llm_chat_server.db
-      - ./config.ini:/app/config.ini
-    environment:
-      - FLASK_ENV=production
-    restart: unless-stopped
-
-volumes:
-  llm_chat_server_db:
--- a/index.html
+++ b/index.html
@ -9,8 +9,6 @@
    <script src="https://cdn.jsdelivr.net/npm/moment@2.29.4/moment.min.js"></script>
    <script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
    <script src="https://cdn.jsdelivr.net/npm/chartjs-adapter-moment@1.0.1/dist/chartjs-adapter-moment.min.js"></script>
-    <script src="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.7.0/highlight.min.js"></script>
-    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/highlight.js/11.7.0/styles/default.min.css">
    <link href="https://fonts.googleapis.com/css2?family=Noto+Sans+Mono:wght@400;700&display=swap" rel="stylesheet">
    <style>
        body {
@ -88,19 +86,13 @@
            color: #888;
        }
        .thought-summary {
-            font-weight: bold;
+            cursor: pointer;
+            color: #fff;
            margin-bottom: 5px;
-            padding: 5px;
-            border-radius: 3px;
+            font-weight: bold;
+            display: flex;
+            align-items: center;
        }
-        .thought-summary.plan { background-color: #2c3e50; }
-        .thought-summary.decision { background-color: #34495e; }
-        .thought-summary.tool_call { background-color: #16a085; }
-        .thought-summary.tool_result { background-color: #27ae60; }
-        .thought-summary.think_more { background-color: #2980b9; }
-        .thought-summary.answer { background-color: #8e44ad; }
-        .thought-summary.reply { background-color: #f39c12; }
-        .thought-summary.thoughts { background-color: #f39c12; }
        .thought-details {
            display: none;
            margin-left: 20px;
@ -246,111 +238,11 @@
                transform: translateX(0);
            }
        }
-
-        .conversation-history-container {
-            margin-top: 20px;
-            background-color: #222;
-            border-radius: 5px;
-            padding: 10px;
-        }
-
-        #conversation-history {
-            color: #fff;
-            font-family: 'Noto Sans Mono', monospace;
-            font-size: 12px;
-        }
-        .history-card {
-            background-color: #2c3e50;
-            border-radius: 5px;
-            padding: 10px;
-            margin-bottom: 10px;
-        }
-        .history-role {
-            font-weight: bold;
-            margin-bottom: 5px;
-        }
-        .history-content {
-            white-space: pre-wrap;
-            word-break: break-word;
-        }
-
-        .error-message {
-            background-color: #ff6b6b;
-            color: #fff;
-            padding: 10px;
-            border-radius: 5px;
-            margin-bottom: 10px;
-        }
-
-        .retrying {
-            background-color: #feca57;
-            color: #333;
-        }
-
-        #clear-history-button {
-            background-color: #e74c3c;
-            color: white;
-            border: none;
-            padding: 10px;
-            margin-bottom: 10px;
-            cursor: pointer;
-            font-family: 'Noto Sans Mono', monospace;
-            font-size: 14px;
-            border-radius: 5px;
-        }
-
-        #clear-history-button:hover {
-            background-color: #c0392b;
-        }
-
-        #chat-tabs {
-            display: flex;
-            background-color: #222;
-            padding: 10px 10px 0 10px;
-        }
-        
-        .chat-tab {
-            background-color: #444;
-            color: #fff;
-            border: none;
-            padding: 10px 20px;
-            margin-right: 5px;
-            cursor: pointer;
-            border-top-left-radius: 5px;
-            border-top-right-radius: 5px;
-        }
-        
-        .chat-tab.active {
-            background-color: #666;
-        }
-        
-        #new-chat-button {
-            background-color: #27ae60;
-            color: #fff;
-            border: none;
-            padding: 10px 20px;
-            cursor: pointer;
-            border-top-left-radius: 5px;
-            border-top-right-radius: 5px;
-        }
-        
-        .close-tab {
-            margin-left: 10px;
-            color: #ff6b6b;
-            cursor: pointer;
-        }
-        
-        .thinking-section {
-            margin-bottom: 20px;
-            border-left: 2px solid #444;
-            padding-left: 10px;
-        }
    </style>
 </head>
 <body>
    <div id="main-container">
        <div id="chat-area">
-            <div id="chat-tabs"></div>
            <div id="chat-container"></div>
            <div id="input-container" class="pdp-panel">
                <div class="pdp-label">INPUT:</div>
@ -380,12 +272,6 @@
                <div class="graph-title">GPU Memory</div>
                <canvas id="gpuMemoryChart"></canvas>
            </div>
-            
-            <!-- Add this new section for conversation history -->
-            <div class="conversation-history-container">
-                <div class="graph-title">Conversation History</div>
-                <div id="conversation-history"></div>
-            </div>
        </div>
    </div>

@ -394,80 +280,10 @@
        const chatContainer = document.getElementById('chat-container');
        const userInput = document.getElementById('user-input');
        const sendButton = document.getElementById('send-button');
-        const chatTabs = document.getElementById('chat-tabs');

-        let currentChatId = null;
-        let chats = {};
-
-        function createNewChat() {
-            const chatId = Date.now().toString();
-            chats[chatId] = {
-                messages: [],
-                thinkingSections: []
-            };
-            addChatTab(chatId);
-            switchToChat(chatId);
-            saveChats();
-        }
-
-        function addChatTab(chatId) {
-            const tab = document.createElement('button');
-            tab.classList.add('chat-tab');
-            tab.textContent = `Chat ${Object.keys(chats).length}`;
-            tab.onclick = () => switchToChat(chatId);
-
-            const closeButton = document.createElement('span');
-            closeButton.classList.add('close-tab');
-            closeButton.textContent = '×';
-            closeButton.onclick = (e) => {
-                e.stopPropagation();
-                closeChat(chatId);
-            };
-
-            tab.appendChild(closeButton);
-            chatTabs.insertBefore(tab, chatTabs.lastElementChild);
-        }
-
-        function switchToChat(chatId) {
-            currentChatId = chatId;
-            document.querySelectorAll('.chat-tab').forEach(tab => tab.classList.remove('active'));
-            document.querySelector(`.chat-tab:nth-child(${Object.keys(chats).indexOf(chatId) + 1})`).classList.add('active');
-            renderChat(chatId);
-        }
-
-        function closeChat(chatId) {
-            delete chats[chatId];
-            saveChats();
-            const tabToRemove = Array.from(chatTabs.children).find(tab => tab.textContent.includes(`Chat ${Object.keys(chats).indexOf(chatId) + 1}`));
-            if (tabToRemove) {
-                chatTabs.removeChild(tabToRemove);
-            }
-            if (currentChatId === chatId) {
-                const remainingChatIds = Object.keys(chats);
-                if (remainingChatIds.length > 0) {
-                    switchToChat(remainingChatIds[0]);
-                } else {
-                    createNewChat();
-                }
-            }
-        }
-
-        function renderChat(chatId) {
-            chatContainer.innerHTML = '';
-            const chat = chats[chatId];
-            chat.messages.forEach(message => addMessage(message.content, message.isUser));
-            chat.thinkingSections.forEach(section => {
-                const thinkingSection = createThinkingSection();
-                section.thoughts.forEach(thought => addThought(thought.type, thought.content, thought.details, thinkingSection));
-            });
-        }
-
-        function createThinkingSection() {
-            const section = document.createElement('div');
-            section.classList.add('thinking-section');
-            chatContainer.appendChild(section);
-            return section;
-        }
+        let thinkingElement = null;
+        let thinkingDetails = null;
+        let thinkingStartTime = null;

        function addMessage(message, isUser) {
            const messageElement = document.createElement('div');
@ -476,40 +292,65 @@
            messageElement.innerHTML = isUser ? message : marked.parse(message);
            chatContainer.appendChild(messageElement);
            chatContainer.scrollTop = chatContainer.scrollHeight;
+        }

-            if (currentChatId) {
-                chats[currentChatId].messages.push({ content: message, isUser: isUser });
-                saveChats();
+        function startThinking() {
+            thinkingElement = document.createElement('div');
+            thinkingElement.classList.add('thought-summary', 'collapsible');
+            
+            const led = document.createElement('div');
+            led.classList.add('led', 'blinking');
+            
+            const textNode = document.createTextNode('Thinking...');
+            
+            thinkingElement.appendChild(led);
+            thinkingElement.appendChild(textNode);
+            thinkingElement.onclick = toggleThinkingDetails;
+            
+            thinkingDetails = document.createElement('div');
+            thinkingDetails.classList.add('thought-details');
+            
+            chatContainer.appendChild(thinkingElement);
+            chatContainer.appendChild(thinkingDetails);
+            
+            thinkingStartTime = Date.now();
+            chatContainer.scrollTop = chatContainer.scrollHeight;
+        }
+
+        function addThought(step, content) {
+            if (thinkingDetails) {
+                const stepElement = document.createElement('div');
+                stepElement.classList.add('thought-summary', 'collapsible');
+                stepElement.textContent = step;
+                stepElement.onclick = toggleStepDetails;
+
+                const stepDetails = document.createElement('div');
+                stepDetails.classList.add('thought-details');
+                stepDetails.innerHTML = content;
+
+                thinkingDetails.appendChild(stepElement);
+                thinkingDetails.appendChild(stepDetails);
+                chatContainer.scrollTop = chatContainer.scrollHeight;
            }
        }

-        function addThought(type, content, details = '', thinkingSection) {
-            const stepElement = document.createElement('div');
-            stepElement.classList.add('thought-summary', 'collapsible', type);
-            stepElement.textContent = type.charAt(0).toUpperCase() + type.slice(1).replace('_', ' ') + ':';
-            stepElement.onclick = toggleStepDetails;
-
-            const stepDetails = document.createElement('div');
-            stepDetails.classList.add('thought-details');
-            
-            if (type === 'error') {
-                stepElement.classList.add('error-message');
-                if (content.includes('retrying')) {
-                    stepElement.classList.add('retrying');
-                }
-                stepDetails.innerHTML = marked.parse(content + '\n\nDetails:\n```\n' + details + '\n```');
-            } else {
-                stepDetails.innerHTML = marked.parse(content);
+        function endThinking(thinkingTime) {
+            if (thinkingElement) {
+                const textNode = thinkingElement.childNodes[1];
+                textNode.nodeValue = `Thinking... (${thinkingTime}s)`;
+                const led = thinkingElement.querySelector('.led');
+                led.classList.remove('blinking');
+                led.style.backgroundColor = '#0f0';
+                led.style.boxShadow = '0 0 10px #0f0';
+                thinkingStartTime = null;
            }
+        }

-            thinkingSection.appendChild(stepElement);
-            thinkingSection.appendChild(stepDetails);
-            chatContainer.scrollTop = chatContainer.scrollHeight;
-
-            if (currentChatId) {
-                const currentThinkingSection = chats[currentChatId].thinkingSections[chats[currentChatId].thinkingSections.length - 1];
-                currentThinkingSection.thoughts.push({ type, content, details });
-                saveChats();
+        function toggleThinkingDetails() {
+            this.classList.toggle('open');
+            const details = this.nextElementSibling;
+            if (details) {
+                details.style.display = details.style.display === 'none' ? 'block' : 'none';
            }
        }

@ -521,71 +362,34 @@
            }
        }

-        function saveChats() {
-            localStorage.setItem('chats', JSON.stringify(chats));
-        }
-
-        function loadChats() {
-            const storedChats = localStorage.getItem('chats');
-            if (storedChats) {
-                chats = JSON.parse(storedChats);
-                Object.keys(chats).forEach(chatId => addChatTab(chatId));
-                if (Object.keys(chats).length > 0) {
-                    switchToChat(Object.keys(chats)[0]);
-                } else {
-                    createNewChat();
-                }
-            } else {
-                createNewChat();
-            }
-        }
-
-        function sendMessage() {
-            const message = userInput.value.trim();
-            if (message && currentChatId) {
-                addMessage(message, true);
-                chats[currentChatId].thinkingSections.push({ thoughts: [] });
-                socket.emit('chat_request', { 
-                    message: message,
-                    conversation_history: chats[currentChatId].messages.filter(m => !m.isUser).map(m => ({ role: 'assistant', content: m.content }))
-                        .concat(chats[currentChatId].messages.filter(m => m.isUser).map(m => ({ role: 'user', content: m.content })))
-                });
-                userInput.value = '';
-            }
-        }
-
        socket.on('thinking', (data) => {
-            if (currentChatId) {
-                const newThinkingSection = createThinkingSection();
-                chats[currentChatId].thinkingSections.push({ thoughts: [] });
-                addThought(data.step, 'Started', '', newThinkingSection);
-            }
+            if (!thinkingElement) startThinking();
+            addThought(data.step, 'Started');
        });

        socket.on('thought', (data) => {
-            if (currentChatId) {
-                const currentThinkingSection = chatContainer.querySelector('.thinking-section:last-child');
-                addThought(data.type, data.content, data.details, currentThinkingSection);
-            }
+            addThought('Result', data.content);
        });

        socket.on('chat_response', (data) => {
-            if (currentChatId) {
-                addMessage(data.response, false);
-            }
+            endThinking(data.thinking_time);
+            addMessage(data.response, false);
        });

        socket.on('error', (data) => {
-            if (currentChatId) {
-                const currentThinkingSection = chatContainer.querySelector('.thinking-section:last-child');
-                if (data.type === 'retrying') {
-                    addThought('error', data.content, '', currentThinkingSection);
-                } else {
-                    addThought('error', data.message, '', currentThinkingSection);
-                }
-            }
+            endThinking(data.thinking_time);
+            addMessage(`Error: ${data.message}`, false);
        });

+        function sendMessage() {
+            const message = userInput.value.trim();
+            if (message) {
+                addMessage(message, true);
+                socket.emit('chat_request', { message: message });
+                userInput.value = '';
+            }
+        }
+
        sendButton.addEventListener('click', sendMessage);
        userInput.addEventListener('keypress', function(e) {
            if (e.key === 'Enter' && !e.shiftKey) {
@ -594,16 +398,6 @@
            }
        });

-        // Add new chat button
-        const newChatButton = document.createElement('button');
-        newChatButton.id = 'new-chat-button';
-        newChatButton.textContent = '+ New Chat';
-        newChatButton.onclick = createNewChat;
-        chatTabs.appendChild(newChatButton);
-
-        // Load chats when the page loads
-        loadChats();
-
        const chartOptions = {
            type: 'line',
            options: {
@ -776,41 +570,6 @@

        window.addEventListener('resize', checkWindowSize);
        checkWindowSize(); // Initial check
-
-        // Add this new function to update the conversation history
-        function updateConversationHistory(history) {
-            const conversationHistoryElement = document.getElementById('conversation-history');
-            conversationHistoryElement.innerHTML = '';
-            
-            history.forEach(item => {
-                const card = document.createElement('div');
-                card.classList.add('history-card');
-                
-                const role = document.createElement('div');
-                role.classList.add('history-role');
-                role.textContent = item.role.charAt(0).toUpperCase() + item.role.slice(1);
-                
-                const content = document.createElement('pre');
-                content.classList.add('history-content');
-                content.innerHTML = hljs.highlightAuto(item.content).value;
-                
-                card.appendChild(role);
-                card.appendChild(content);
-                conversationHistoryElement.appendChild(card);
-            });
-        }
-
-        // Add this new socket listener
-        socket.on('conversation_history', (data) => {
-            updateConversationHistory(data.history);
-        });
-
-        // Add event listener for the clear history button
-        clearHistoryButton.addEventListener('click', () => {
-            if (confirm('Are you sure you want to clear the conversation history?')) {
-                clearConversationHistory();
-            }
-        });
    </script>
 </body>
 </html>
--- a/main.py
+++ b/main.py
--- a/models.py
+++ b/models.py
@ -3,86 +3,30 @@ import structlog

 logger = structlog.get_logger()

-
 class ModelManager:
    def __init__(self):
        self.model_capabilities = {
-            "ajindal/llama3.1-storm:8b": [
-                "general_knowledge",
-                "reasoning",
-                "tool_calling",
-                "conversation",
-                "multilingual",
-                "instruction_following",
-            ],
-            "llama3.1:8b": [
-                "general_knowledge",
-                "reasoning",
-                "tool_calling",
-                "conversation",
-                "multilingual",
-                "instruction_following",
-            ],
-            "qwen2.5:7b": [
-                "general_knowledge",
-                "reasoning",
-                "tool_calling",
-                "conversation",
-                "multilingual",
-                "instruction_following",
-            ],
-            "llama3.2:3b": [
-                "summarization",
-                "instruction_following",
-                "tool_calling",
-                "multilingual",
-            ],
-            "llava:7b": [
-                "visual_reasoning",
-                "visual_conversation",
-                "visual_tool_calling",
-                "vision",
-                "ocr",
-                "multimodal",
-            ],
+            "qwen2.5:7b": ["general_knowledge", "structured_output", "multilingual", "instruction_following", "structured_data"],
+            "llama3.1:8b": ["general_knowledge", "reasoning", "tool_calling", "conversation", "multilingual", "instruction_following"],
+            "qwen2.5-coder:7b": ["code_generation", "code_analysis", "instruction_following", "math_reasoning"],
+            "llama3.2:3b": ["summarization", "instruction_following", "tool_calling", "multilingual"],
+            "llava:7b": ["visual_reasoning", "visual_conversation", "visual_tool_calling", "vision", "ocr", "multimodal"],
        }
-        logger.info(
-            "ModelManager initialized", model_capabilities=self.model_capabilities
-        )
+        logger.info("ModelManager initialized", model_capabilities=self.model_capabilities)

    def get_model_capabilities(self, model_name):
        capabilities = self.model_capabilities.get(model_name, [])
-        logger.debug(
-            "Retrieved model capabilities", model=model_name, capabilities=capabilities
-        )
+        logger.debug("Retrieved model capabilities", model=model_name, capabilities=capabilities)
        return capabilities

    def select_best_model(self, required_capability):
-        suitable_models = [
-            model
-            for model, capabilities in self.model_capabilities.items()
-            if required_capability in capabilities
-        ]
-        selected_model = (
-            suitable_models[0]
-            if suitable_models
-            else list(self.model_capabilities.keys())[0]
-        )
-        logger.info(
-            "Selected best model",
-            required_capability=required_capability,
-            selected_model=selected_model,
-        )
+        suitable_models = [model for model, capabilities in self.model_capabilities.items() if required_capability in capabilities]
+        selected_model = suitable_models[0] if suitable_models else list(self.model_capabilities.keys())[0]
+        logger.info("Selected best model", required_capability=required_capability, selected_model=selected_model)
        return selected_model

-    def generate_text(
-        self,
-        model_name,
-        prompt,
-        max_length=100,
-        system="You are a helpful assistant.",
-        tools=[],
-    ):
+    def generate_text(self, model_name, prompt, max_length=100, system="You are a helpful assistant.", stream=False):
+        logger.debug("Generating text", model=model_name, prompt=prompt, max_length=max_length)
        # Check if model exists
        try:
            ollama.pull(model_name)
@ -94,16 +38,8 @@ class ModelManager:
            else:
                logger.exception("Error pulling model", model=model_name, error=str(e))
                raise e
+        response = ollama.generate(model=model_name, prompt=prompt, system=system, stream=stream)
+        logger.debug("Text generated", model=model_name, response=response['response'])
+        return response['response']

-        response = ollama.generate(
-            model=model_name,
-            prompt=prompt,
-            system=system,
-            tools=tools,
-            max_tokens=max_length,
-        )
-        logger.debug("Text generated", model=model_name, response=response["response"])
-        return response["response"]
-
-
-model_manager = ModelManager()
+model_manager = ModelManager()
--- a/requirements.txt
+++ b/requirements.txt
@ -4,19 +4,11 @@ aiohttp==3.10.5
 aiosignal==1.3.1
 annotated-types==0.7.0
 anyio==4.6.0
-art==6.3
 attrs==24.2.0
-beautifulsoup4==4.12.3
-bidict==0.23.1
-black==24.8.0
-blinker==1.8.2
-bs4==0.0.2
 certifi==2024.7.4
-chardet==5.2.0
 charset-normalizer==3.3.2
 click==8.1.7
 cloudpickle==3.0.0
-cssselect==1.2.0
 datasets==3.0.0
 dill==0.3.8
 diskcache==5.6.3
@ -25,13 +17,9 @@ duckduckgo_search==6.2.6
 einops==0.8.0
 fastapi==0.115.0
 filelock==3.15.4
-Flask==3.0.3
-flask-openapi3==3.1.3
-Flask-SocketIO==5.3.7
 frozenlist==1.4.1
 fsspec==2024.6.1
 gguf==0.9.1
-GPUtil==1.4.0
 h11==0.14.0
 httpcore==1.0.5
 httptools==0.6.1
@ -41,8 +29,6 @@ idna==3.7
 importlib_metadata==8.5.0
 inquirerpy==0.3.4
 interegular==0.3.3
-isort==5.13.2
-itsdangerous==2.2.0
 Jinja2==3.1.4
 jiter==0.5.0
 jsonschema==4.23.0
@ -50,9 +36,6 @@ jsonschema-specifications==2023.12.1
 lark==1.2.2
 llvmlite==0.43.0
 lm-format-enforcer==0.10.6
-lxml==5.3.0
-lxml_html_clean==0.2.2
-markdownify==0.13.1
 MarkupSafe==2.1.5
 mistral_common==1.4.3
 mpmath==1.3.0
@ -60,7 +43,6 @@ msgpack==1.1.0
 msgspec==0.18.6
 multidict==6.1.0
 multiprocess==0.70.16
-mypy-extensions==1.0.0
 nest-asyncio==1.6.0
 networkx==3.3
 numba==0.60.0
@ -78,16 +60,13 @@ nvidia-ml-py==12.560.30
 nvidia-nccl-cu12==2.20.5
 nvidia-nvjitlink-cu12==12.6.20
 nvidia-nvtx-cu12==12.1.105
-ollama==0.3.3
 openai==1.47.1
 outlines==0.0.46
 packaging==24.1
 pandas==2.2.3
 partial-json-parser==0.2.1.1.post4
-pathspec==0.12.1
 pfzy==0.3.4
 pillow==10.4.0
-platformdirs==4.3.6
 primp==0.5.5
 prometheus-fastapi-instrumentator==7.0.0
 prometheus_client==0.21.0
@ -102,13 +81,10 @@ pydantic==2.9.2
 pydantic_core==2.23.4
 python-dateutil==2.9.0.post0
 python-dotenv==1.0.1
-python-engineio==4.9.1
-python-socketio==5.11.4
 pytz==2024.2
 PyYAML==6.0.2
 pyzmq==26.2.0
 ray==2.36.1
-readability-lxml==0.8.1
 referencing==0.35.1
 regex==2024.7.24
 requests==2.32.3
@ -116,12 +92,9 @@ rpds-py==0.20.0
 safetensors==0.4.4
 sentencepiece==0.2.0
 setuptools==72.1.0
-simple-websocket==1.0.0
 six==1.16.0
 sniffio==1.3.1
-soupsieve==2.6
 starlette==0.38.6
-structlog==24.4.0
 sympy==1.13.2
 tiktoken==0.7.0
 tokenizers==0.19.1
@ -140,8 +113,6 @@ vllm-flash-attn==2.6.1
 watchfiles==0.24.0
 wcwidth==0.2.13
 websockets==13.1
-Werkzeug==3.0.4
-wsproto==1.2.0
 xformers==0.0.27.post2
 xxhash==3.5.0
 yarl==1.12.0
--- a/schema.sql
+++ b/schema.sql
@ -1,16 +0,0 @@
-CREATE TABLE IF NOT EXISTS Keys (
-    id INTEGER PRIMARY KEY AUTOINCREMENT,
-    username TEXT NOT NULL UNIQUE,
-    api_key TEXT NOT NULL UNIQUE
-);
-
-CREATE TABLE IF NOT EXISTS Queries (
-    id TEXT PRIMARY KEY,
-    ip TEXT NOT NULL,
-    timestamp DATETIME DEFAULT CURRENT_TIMESTAMP,
-    query TEXT NOT NULL,
-    api_key_id INTEGER,
-    status TEXT NOT NULL,
-    conversation_history TEXT,
-    FOREIGN KEY (api_key_id) REFERENCES Keys (id)
-);
--- a/tools.py
+++ b/tools.py
@ -1,29 +1,16 @@
-import subprocess
-import tempfile
-import time
-import json
-import requests
-from markdownify import markdownify as md
-from readability.readability import Document
 import duckduckgo_search
-import datetime
-import random
-import math
-import re
-import base64
-from io import BytesIO
-from PIL import Image, ImageDraw, ImageFont
-import ollama
-import os
+import requests
+from readability.readability import Document
+from markdownify import markdownify as md

 class Tool:
-    def __init__(self, name: str, description: str, arguments: dict, returns: str):
+    def __init__(self, name: str, description: str, arguments: str, returns: str):
        self.name = name
        self.description = description
        self.arguments = arguments
        self.returns = returns

-    def execute(self, arguments: dict) -> str:
+    def execute(self, arguments: str) -> str:
        pass


@ -39,23 +26,10 @@ class ToolManager:
            if tool.name == name:
                return tool
        return None
-
+    
    def get_tools_and_descriptions_for_prompt(self):
        return "\n".join([f"{tool.name}: {tool.description}" for tool in self.tools])
-
-    def get_tools_for_ollama_dict(self):
-        return [
-            {
-                "type": "function",
-                "function": {
-                    "name": tool.name,
-                    "description": tool.description,
-                    "parameters": tool.arguments,
-                },
-            }
-            for tool in self.tools
-        ]
-
+    

 class DefaultToolManager(ToolManager):
    def __init__(self):
@ -64,35 +38,16 @@ class DefaultToolManager(ToolManager):
        self.add_tool(GetReadablePageContentsTool())
        self.add_tool(CalculatorTool())
        self.add_tool(PythonCodeTool())
-        self.add_tool(DateTimeTool())
-        self.add_tool(RandomNumberTool())
-        self.add_tool(RegexTool())
-        self.add_tool(Base64Tool())
-        self.add_tool(SimpleChartTool())
-        self.add_tool(LLAVAImageAnalysisTool())


 class SearchTool(Tool):
    def __init__(self):
-        super().__init__(
-            "search_web",
-            "Search the internet for information",
-            {
-                "type": "object",
-                "properties": {
-                    "query": {"type": "string", "description": "The search query"}
-                },
-            },
-            "results:list[string]",
-        )
-
-    def execute(self, arg: dict) -> str:
-        try:
-            res = duckduckgo_search.DDGS().text(arg["query"], max_results=5)
-            return "\n\n".join([f"{r['title']}\n{r['body']}\n{r['href']}" for r in res])
-        except Exception as e:
-            return f"Error searching the web: {str(e)}"
+        super().__init__("search_web", "Search the internet for information -- Takes a search query as an argument", "query:string", "results:list[string]")

+    def execute(self, arg: str) -> str:
+        res = duckduckgo_search.DDGS().text(arg, max_results=5)
+        return [f"{r['title']}\n{r['body']}\n{r['href']}" for r in res]
+    

 def get_readable_page_contents(url: str) -> str:
    try:
@ -103,267 +58,28 @@ def get_readable_page_contents(url: str) -> str:
        return md(content)
    except Exception as e:
        return f"Error fetching readable content: {str(e)}"
+    


 class GetReadablePageContentsTool(Tool):
    def __init__(self):
-        super().__init__(
-            "get_readable_page_contents",
-            "Get the contents of a web page in a readable format",
-            {
-                "type": "object",
-                "properties": {
-                    "url": {"type": "string", "description": "The url of the web page"}
-                },
-            },
-            "contents:string",
-        )
+        super().__init__("get_readable_page_contents", "Get the contents of a web page in a readable format -- Takes a url as an argument", "url:string", "contents:string")

-    def execute(self, arg: dict) -> str:
-        return get_readable_page_contents(arg["url"])
+    def execute(self, arg: str) -> str:
+        return get_readable_page_contents(arg[0])


 class CalculatorTool(Tool):
    def __init__(self):
-        super().__init__(
-            "calculator",
-            "Perform a calculation using python's eval function",
-            {
-                "type": "object",
-                "properties": {
-                    "expression": {
-                        "type": "string",
-                        "description": "The mathematical expression to evaluate, should be a python mathematical expression",
-                    }
-                },
-            },
-            "result:string",
-        )
+        super().__init__("calculator", "Perform a calculation -- Takes a python mathematical expression as an argument", "expression:string", "result:string")

-    def execute(self, arg: dict) -> str:
-        try:
-            return str(eval(arg["expression"]))
-        except Exception as e:
-            return f"Error executing code: {str(e)}"
+    def execute(self, arg: str) -> str:
+        return str(eval(arg[0]))


 class PythonCodeTool(Tool):
    def __init__(self):
-        super().__init__(
-            "python_code",
-            "Execute python code using a temporary file and a subprocess. You must print results to stdout.",
-            {
-                "type": "object",
-                "properties": {
-                    "code": {
-                        "type": "string",
-                        "description": "The python code to execute, can be multiline",
-                    }
-                },
-            },
-            "result:string",
-        )
+        super().__init__("python_code", "Execute python code -- Takes a python code as an argument, code must be a single line of valid python", "code:string", "result:string")

-    def execute(self, arg: dict) -> str:
-        try:
-            with tempfile.NamedTemporaryFile(
-                suffix=".py", mode="w", delete=False
-            ) as temp_file:
-                temp_file.write(arg["code"])
-                temp_file.flush()
-
-                start_time = time.time()
-                process = subprocess.Popen(
-                    ["python", temp_file.name],
-                    stdout=subprocess.PIPE,
-                    stderr=subprocess.PIPE,
-                    text=True,
-                )
-                stdout, stderr = process.communicate(timeout=10)  # 10 second timeout
-                end_time = time.time()
-                execution_time = end_time - start_time
-
-                result = {
-                    "stdout": stdout,
-                    "stderr": stderr,
-                    "return_value": process.returncode,
-                    "execution_time": execution_time,
-                }
-
-        except subprocess.TimeoutExpired:
-            process.kill()
-            return "Error: Code execution timed out after 10 seconds"
-        except Exception as e:
-            return f"Error executing code: {str(e)}"
-
-        return "\n".join([f"{k}:\n{v}" for k, v in result.items()])
-
-
-class DateTimeTool(Tool):
-    def __init__(self):
-        super().__init__(
-            "get_current_datetime",
-            "Get the current date and time",
-            {"type": "object", "properties": {}},
-            "datetime:string"
-        )
-
-    def execute(self, arg: dict) -> str:
-        return datetime.datetime.now().strftime("%Y-%m-%d %H:%M:%S")
-
-
-class RandomNumberTool(Tool):
-    def __init__(self):
-        super().__init__(
-            "generate_random_number",
-            "Generate a random number within a given range",
-            {
-                "type": "object",
-                "properties": {
-                    "min": {"type": "number", "description": "The minimum value"},
-                    "max": {"type": "number", "description": "The maximum value"}
-                }
-            },
-            "random_number:number"
-        )
-
-    def execute(self, arg: dict) -> str:
-        return str(random.uniform(arg["min"], arg["max"]))
-
-
-class RegexTool(Tool):
-    def __init__(self):
-        super().__init__(
-            "regex_match",
-            "Perform a regex match on a given text",
-            {
-                "type": "object",
-                "properties": {
-                    "text": {"type": "string", "description": "The text to search in"},
-                    "pattern": {"type": "string", "description": "The regex pattern to match"}
-                }
-            },
-            "matches:list[string]"
-        )
-
-    def execute(self, arg: dict) -> str:
-        matches = re.findall(arg["pattern"], arg["text"])
-        return json.dumps(matches)
-
-
-class Base64Tool(Tool):
-    def __init__(self):
-        super().__init__(
-            "base64_encode_decode",
-            "Encode or decode a string using Base64",
-            {
-                "type": "object",
-                "properties": {
-                    "action": {"type": "string", "enum": ["encode", "decode"], "description": "Whether to encode or decode"},
-                    "text": {"type": "string", "description": "The text to encode or decode"}
-                }
-            },
-            "result:string"
-        )
-
-    def execute(self, arg: dict) -> str:
-        if arg["action"] == "encode":
-            return base64.b64encode(arg["text"].encode()).decode()
-        elif arg["action"] == "decode":
-            return base64.b64decode(arg["text"].encode()).decode()
-        else:
-            return "Invalid action. Use 'encode' or 'decode'."
-
-
-class SimpleChartTool(Tool):
-    def __init__(self):
-        super().__init__(
-            "generate_simple_chart",
-            "Generate a simple bar chart image",
-            {
-                "type": "object",
-                "properties": {
-                    "data": {"type": "array", "items": {"type": "number"}, "description": "List of numerical values for the chart"},
-                    "labels": {"type": "array", "items": {"type": "string"}, "description": "Labels for each bar"}
-                }
-            },
-            "image_base64:string"
-        )
-
-    def execute(self, arg: dict) -> str:
-        data = arg["data"]
-        labels = arg["labels"]
-        
-        # Create a simple bar chart
-        width, height = 400, 300
-        img = Image.new('RGB', (width, height), color='white')
-        draw = ImageDraw.Draw(img)
-        
-        # Draw bars
-        max_value = max(data)
-        bar_width = width // (len(data) + 1)
-        for i, value in enumerate(data):
-            bar_height = (value / max_value) * (height - 50)
-            left = (i + 1) * bar_width
-            draw.rectangle([left, height - bar_height, left + bar_width, height], fill='blue')
-        
-        # Add labels
-        font = ImageFont.load_default()
-        for i, label in enumerate(labels):
-            left = (i + 1) * bar_width + bar_width // 2
-            draw.text((left, height - 20), label, fill='black', anchor='ms', font=font)
-        
-        # Convert to base64
-        buffered = BytesIO()
-        img.save(buffered, format="PNG")
-        img_str = base64.b64encode(buffered.getvalue()).decode()
-        return img_str
-
-
-class LLAVAImageAnalysisTool(Tool):
-    def __init__(self):
-        super().__init__(
-            "analyze_image",
-            "Analyze an image using the LLAVA model",
-            {
-                "type": "object",
-                "properties": {
-                    "image_base64": {"type": "string", "description": "Base64 encoded image"},
-                    "question": {"type": "string", "description": "Question about the image"}
-                }
-            },
-            "analysis:string"
-        )
-
-    def execute(self, arg: dict) -> str:
-        try:
-            # Decode base64 image
-            image_data = base64.b64decode(arg["image_base64"])
-            image = Image.open(BytesIO(image_data))
-
-            # Save image to a temporary file
-            with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as temp_file:
-                image.save(temp_file, format="PNG")
-                temp_file_path = temp_file.name
-
-            # Call LLAVA model
-            response = ollama.chat(
-                model="llava:7b",
-                messages=[
-                    {
-                        "role": "user",
-                        "content": arg["question"],
-                        "images": [temp_file_path]
-                    }
-                ]
-            )
-
-            # Clean up temporary file
-            os.remove(temp_file_path)
-
-            # Unload LLAVA model
-            ollama.delete("llava:7b")
-
-            return response['message']['content']
-        except Exception as e:
-            return f"Error analyzing image: {str(e)}"
+    def execute(self, arg: str) -> str:
+        return str(eval(arg[0]))