Initial submission

.gitignore

@@ -0,0 +1,9 @@
+.env
+local_test
+example_usage
+test_agent
+test_all_tools
+test_math_tools
+test_multimodal_tools
+test_search_tools
+test_youtube_tools

app.py

@@ -3,23 +3,86 @@ import gradio as gr
 import requests
 import inspect
 import pandas as pd
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+# Import your LangGraph agent
+from graph.graph_builder import graph
+from langchain_core.messages import HumanMessage
 
 # (Keep Constants as is)
 # --- Constants ---
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
-# --- Basic Agent Definition ---
-# ----- THIS IS WERE YOU CAN BUILD WHAT YOU WANT ------
+# --- Your LangGraph Agent Definition ---
+# ----- THIS IS WHERE YOU BUILD YOUR AGENT ------
 class BasicAgent:
     def __init__(self):
-        print("BasicAgent initialized.")
+        """Initialize the LangGraph agent"""
+        print("LangGraph Agent initialized with multimodal, search, math, and YouTube tools.")
+        
+        # Verify environment variables
+        if not os.getenv("OPENROUTER_API_KEY"):
+            raise ValueError("OPENROUTER_API_KEY not found in environment variables")
+        
+        # The graph is already compiled and ready to use
+        self.graph = graph
+        print("✅ Agent ready with tools: multimodal, search, math, YouTube")
+    
     def __call__(self, question: str) -> str:
-        print(f"Agent received question (first 50 chars): {question[:50]}...")
-        fixed_answer = "This is a default answer."
-        print(f"Agent returning fixed answer: {fixed_answer}")
-        return fixed_answer
+        """
+        Process a question using the LangGraph agent and return just the answer
+        
+        Args:
+            question: The question to answer
+            
+        Returns:
+            str: The final answer (formatted for evaluation)
+        """
+        print(f"🤖 Processing question: {question[:50]}...")
+        
+        try:
+            # Create initial state with the question
+            initial_state = {"messages": [HumanMessage(content=question)]}
+            
+            # Run the LangGraph agent
+            result = self.graph.invoke(initial_state)
+            
+            # Extract the final message content
+            final_message = result["messages"][-1]
+            answer = final_message.content
+            
+            # Clean up the answer for evaluation (remove any extra formatting)
+            # The evaluation system expects just the answer, no explanations
+            if isinstance(answer, str):
+                answer = answer.strip()
+                
+                # Remove common prefixes that might interfere with evaluation
+                prefixes_to_remove = [
+                    "The answer is: ",
+                    "Answer: ",
+                    "The result is: ",
+                    "Result: ",
+                    "The final answer is: ",
+                ]
+                
+                for prefix in prefixes_to_remove:
+                    if answer.startswith(prefix):
+                        answer = answer[len(prefix):].strip()
+                        break
+            
+            print(f"✅ Agent answer: {answer}")
+            return answer
+            
+        except Exception as e:
+            error_msg = f"Error processing question: {str(e)}"
+            print(f"❌ {error_msg}")
+            return error_msg
 
-def run_and_submit_all( profile: gr.OAuthProfile | None):
+# Keep the rest of the file unchanged (run_and_submit_all function and Gradio interface)
+def run_and_submit_all(profile: gr.OAuthProfile | None):
     """
     Fetches all questions, runs the BasicAgent on them, submits all answers,
     and displays the results.
@@ -38,13 +101,14 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
     questions_url = f"{api_url}/questions"
     submit_url = f"{api_url}/submit"
 
-    # 1. Instantiate Agent ( modify this part to create your agent)
+    # 1. Instantiate Agent (using your LangGraph agent)
     try:
         agent = BasicAgent()
     except Exception as e:
         print(f"Error instantiating agent: {e}")
         return f"Error initializing agent: {e}", None
-    # In the case of an app running as a hugging Face space, this link points toward your codebase ( usefull for others so please keep it public)
+    
+    # In the case of an app running as a hugging Face space, this link points toward your codebase
     agent_code = f"https://huggingface.co/spaces/{space_id}/tree/main"
     print(agent_code)
 
@@ -61,10 +125,6 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
     except requests.exceptions.RequestException as e:
         print(f"Error fetching questions: {e}")
         return f"Error fetching questions: {e}", None
-    except requests.exceptions.JSONDecodeError as e:
-         print(f"Error decoding JSON response from questions endpoint: {e}")
-         print(f"Response text: {response.text[:500]}")
-         return f"Error decoding server response for questions: {e}", None
     except Exception as e:
         print(f"An unexpected error occurred fetching questions: {e}")
         return f"An unexpected error occurred fetching questions: {e}", None
@@ -139,22 +199,26 @@ def run_and_submit_all( profile: gr.OAuthProfile | None):
         results_df = pd.DataFrame(results_log)
         return status_message, results_df
 
-
 # --- Build Gradio Interface using Blocks ---
 with gr.Blocks() as demo:
-    gr.Markdown("# Basic Agent Evaluation Runner")
+    gr.Markdown("# LangGraph Agent Evaluation Runner")
     gr.Markdown(
         """
         **Instructions:**
 
-        1.  Please clone this space, then modify the code to define your agent's logic, the tools, the necessary packages, etc ...
-        2.  Log in to your Hugging Face account using the button below. This uses your HF username for submission.
-        3.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
+        This space uses a LangGraph agent with multimodal, search, math, and YouTube tools powered by OpenRouter.
+        
+        1.  Log in to your Hugging Face account using the button below.
+        2.  Click 'Run Evaluation & Submit All Answers' to fetch questions, run your agent, submit answers, and see the score.
 
+        **Agent Capabilities:**
+        - 🎨 **Multimodal**: Analyze images, extract text (OCR), process audio transcripts
+        - 🔍 **Search**: Web search using multiple providers (DuckDuckGo, Tavily, SerpAPI)
+        - 🧮 **Math**: Basic arithmetic, complex calculations, percentages, factorials
+        - 📺 **YouTube**: Extract captions, get video information
+        
         ---
-        **Disclaimers:**
-        Once clicking on the "submit button, it can take quite some time ( this is the time for the agent to go through all the questions).
-        This space provides a basic setup and is intentionally sub-optimal to encourage you to develop your own, more robust solution. For instance for the delay process of the submit button, a solution could be to cache the answers and submit in a seperate action or even to answer the questions in async.
+        **Note:** Processing all questions may take some time as the agent carefully analyzes each question and uses appropriate tools.
         """
     )
 
@@ -163,7 +227,6 @@ with gr.Blocks() as demo:
     run_button = gr.Button("Run Evaluation & Submit All Answers")
 
     status_output = gr.Textbox(label="Run Status / Submission Result", lines=5, interactive=False)
-    # Removed max_rows=10 from DataFrame constructor
     results_table = gr.DataFrame(label="Questions and Agent Answers", wrap=True)
 
     run_button.click(
@@ -175,7 +238,7 @@ if __name__ == "__main__":
     print("\n" + "-"*30 + " App Starting " + "-"*30)
     # Check for SPACE_HOST and SPACE_ID at startup for information
     space_host_startup = os.getenv("SPACE_HOST")
-    space_id_startup = os.getenv("SPACE_ID") # Get SPACE_ID at startup
+    space_id_startup = os.getenv("SPACE_ID")
 
     if space_host_startup:
         print(f"✅ SPACE_HOST found: {space_host_startup}")
@@ -183,7 +246,7 @@ if __name__ == "__main__":
     else:
         print("ℹ️  SPACE_HOST environment variable not found (running locally?).")
 
-    if space_id_startup: # Print repo URLs if SPACE_ID is found
+    if space_id_startup:
         print(f"✅ SPACE_ID found: {space_id_startup}")
         print(f"   Repo URL: https://huggingface.co/spaces/{space_id_startup}")
         print(f"   Repo Tree URL: https://huggingface.co/spaces/{space_id_startup}/tree/main")
@@ -192,5 +255,5 @@ if __name__ == "__main__":
 
     print("-"*(60 + len(" App Starting ")) + "\n")
 
-    print("Launching Gradio Interface for Basic Agent Evaluation...")
-    demo.launch(debug=True, share=False)
+    print("Launching Gradio Interface for LangGraph Agent Evaluation...")
+    demo.launch(debug=True, share=False)

example_usage.py

@@ -0,0 +1,14 @@
+# example_usage.py
+from graph.graph_builder import graph
+from langchain_core.messages import HumanMessage
+
+# Simple example
+def ask_agent(question: str):
+    initial_state = {"messages": [HumanMessage(content=question)]}
+    result = graph.invoke(initial_state)
+    return result["messages"][-1].content
+
+# Usage
+if __name__ == "__main__":
+    response = ask_agent("How many albums did Michael Jackson release in his lifetime?")
+    print(f"Agent: {response}")

graph/graph_builder.py

@@ -0,0 +1,24 @@
+# graph/graph_builder.py (unchanged)
+from langgraph.graph import START, StateGraph
+from langgraph.prebuilt import tools_condition
+from langgraph.prebuilt import ToolNode
+from nodes.core import assistant, tools
+from states.state import AgentState
+
+## The graph
+builder = StateGraph(AgentState)
+
+# Define nodes: these do the work
+builder.add_node("assistant", assistant)
+builder.add_node("tools", ToolNode(tools))
+
+# Define edges: these determine how the control flow moves
+builder.add_edge(START, "assistant")
+builder.add_conditional_edges(
+    "assistant",
+    # If the latest message requires a tool, route to tools
+    # Otherwise, provide a direct response
+    tools_condition,
+)
+builder.add_edge("tools", "assistant")
+graph = builder.compile()

local_test.py

@@ -0,0 +1,17 @@
+# local_test.py
+from app import BasicAgent
+
+agent = BasicAgent()
+
+# Test questions
+test_questions = [
+    "What is 15 + 27?",
+    "Calculate the square root of 144",
+    "What is 25% of 200?"
+]
+
+for q in test_questions:
+    answer = agent(q)
+    print(f"Q: {q}")
+    print(f"A: {answer}")
+    print("-" * 30)

nodes/core.py

@@ -0,0 +1,90 @@
+# nodes/core.py
+from states.state import AgentState
+import os
+from dotenv import load_dotenv
+from langchain_openai import ChatOpenAI  # Using OpenAI-compatible API for OpenRouter
+from tools.langchain_tools import (
+    extract_text, 
+    analyze_image_tool, 
+    analyze_audio_tool,
+    add, 
+    subtract, 
+    multiply, 
+    divide,
+    search_tool,
+    extract_youtube_transcript,
+    get_youtube_info,
+    calculate_expression,
+    factorial,
+    square_root,
+    percentage,
+    average
+)
+
+load_dotenv()
+
+# Read your API key from the environment variable
+openrouter_api_key = os.getenv("OPENROUTER_API_KEY")
+
+if not openrouter_api_key:
+    raise ValueError("OPENROUTER_API_KEY not found in environment variables")
+
+# Initialize OpenRouter ChatOpenAI with OpenRouter-specific configuration
+chat = ChatOpenAI(
+    model="google/gemini-2.5-flash-preview-05-20",  # Free multimodal model
+    # Alternative models you can use:
+    # model="mistralai/mistral-7b-instruct:free",  # Fast, free text model
+    # model="google/gemma-2-9b-it:free",  # Google's free model
+    # model="qwen/qwen-2.5-72b-instruct:free",  # High-quality free model
+    
+    temperature=0,
+    max_retries=2,
+    base_url="https://openrouter.ai/api/v1",
+    api_key=openrouter_api_key,
+    default_headers={
+        "HTTP-Referer": "https://your-app.com",  # Optional: for analytics
+        "X-Title": "LangGraph Agent",  # Optional: for analytics
+    }
+)
+
+# Core tools list (matching original structure)
+tools = [
+    extract_text,
+    analyze_image_tool,
+    analyze_audio_tool,
+    extract_youtube_transcript,
+    add,
+    subtract,
+    multiply,
+    divide,
+    search_tool
+]
+
+# Extended tools list (if you want more capabilities)
+extended_tools = tools + [
+    get_youtube_info,
+    calculate_expression,
+    factorial,
+    square_root,
+    percentage,
+    average
+]
+
+# Use core tools by default (matching original), but you can switch to extended_tools
+chat_with_tools = chat.bind_tools(tools)
+
+def assistant(state: AgentState):
+    """
+    Assistant node - maintains the exact same system prompt for evaluation compatibility
+    """
+    sys_msg = (
+        "You are a helpful assistant with access to tools. Understand user requests accurately. "
+        "Use your tools when needed to answer effectively. Strictly follow all user instructions and constraints. "
+        "Pay attention: your output needs to contain only the final answer without any reasoning since it will be "
+        "strictly evaluated against a dataset which contains only the specific response. "
+        "Your final output needs to be just the string or integer containing the answer, not an array or technical stuff."
+    )
+    
+    return {
+        "messages": [chat_with_tools.invoke([sys_msg] + state["messages"])]
+    }

requirements.txt

@@ -1,2 +1,13 @@
+python-dotenv
+requests
+pytubefix
+pillow
+langgraph
+langchain
+langchain-openai
+langchain-core
+langchain-community
 gradio
-requests
+pandas
+gradio[oauth]
+

states/state.py

@@ -0,0 +1,7 @@
+# states/state.py (unchanged)
+from typing import TypedDict, Annotated
+from langchain_core.messages import AnyMessage
+from langgraph.graph.message import add_messages
+
+class AgentState(TypedDict):
+    messages: Annotated[list[AnyMessage], add_messages]

test_agent.py

@@ -0,0 +1,50 @@
+# test_agent.py
+import os
+from dotenv import load_dotenv
+from graph.graph_builder import graph
+from langchain_core.messages import HumanMessage
+
+load_dotenv()
+
+def test_agent():
+    """Test the LangGraph agent with various queries"""
+    
+    # Check if OpenRouter API key is available
+    if not os.getenv("OPENROUTER_API_KEY"):
+        print("❌ OPENROUTER_API_KEY not found in environment")
+        return
+    
+    print("🤖 Testing LangGraph Agent with OpenRouter")
+    print("=" * 50)
+    
+    # Test cases
+    test_queries = [
+        "What is 15 + 27?",
+        "Calculate the square root of 144",
+        "Search for information about artificial intelligence",
+        "What is 25% of 200?"
+    ]
+    
+    for i, query in enumerate(test_queries, 1):
+        print(f"\n🧪 Test {i}: {query}")
+        print("-" * 30)
+        
+        try:
+            # Create initial state
+            initial_state = {"messages": [HumanMessage(content=query)]}
+            
+            # Run the agent
+            result = graph.invoke(initial_state)
+            
+            # Get the final message
+            final_message = result["messages"][-1]
+            print(f"🤖 Response: {final_message.content}")
+            
+        except Exception as e:
+            print(f"❌ Error: {str(e)}")
+    
+    print("\n" + "=" * 50)
+    print("✅ Agent testing completed!")
+
+if __name__ == "__main__":
+    test_agent()

test_all_tools.py

@@ -0,0 +1,63 @@
+# test_all_tools.py
+import os
+import sys
+from dotenv import load_dotenv
+from pathlib import Path
+
+# Load environment variables
+load_dotenv()
+
+# Add tools directory to path
+sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
+
+def main():
+    """Run all tool tests"""
+    print("=" * 60)
+    print("🧪 TESTING ALL TOOLS")
+    print("=" * 60)
+    
+    # Check if OpenRouter API key is loaded
+    openrouter_key = os.getenv("OPENROUTER_API_KEY")
+    if openrouter_key:
+        print(f"✅ OpenRouter API Key loaded: {openrouter_key[:8]}...")
+    else:
+        print("❌ OpenRouter API Key not found in environment")
+        print("Please add OPENROUTER_API_KEY to your .env file")
+    
+    # Import and run individual test modules
+    try:
+        print("\n🔧 Testing Math Tools...")
+        from test_math_tools import test_math_tools
+        test_math_tools()
+        
+        print("\n🔍 Testing Search Tools...")
+        from test_search_tools import test_search_tools
+        test_search_tools()
+        
+        print("\n🎨 Testing Multimodal Tools...")
+        from test_multimodal_tools import test_multimodal_tools
+        test_multimodal_tools()
+        
+        print("\n📺 Testing YouTube Tools...")
+        from test_youtube_tools import test_youtube_tools
+        test_youtube_tools()
+        
+        print("\n" + "=" * 60)
+        print("✅ ALL TESTS COMPLETED SUCCESSFULLY!")
+        print("=" * 60)
+        
+        # Usage examples
+        print("\n📚 Quick Usage Examples:")
+        print("from tools import get_video_info, search_web, add, analyze_image")
+        print("video_info = get_video_info('https://youtube.com/watch?v=...')")
+        print("search_results = search_web('Python programming')")
+        print("result = add(10, 5)")
+        print("image_desc = analyze_image('photo.jpg', 'What is in this image?')")
+        
+    except Exception as e:
+        print(f"\n❌ Error running tests: {str(e)}")
+        import traceback
+        traceback.print_exc()
+
+if __name__ == "__main__":
+    main()

test_math_tools.py

@@ -0,0 +1,78 @@
+# test_math_tools.py
+import sys
+import os
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+# Add tools to path
+sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
+
+def test_math_tools():
+    """Test all math tool functions"""
+    try:
+        from tools.math_tools import MathTools, add, subtract, multiply, divide, factorial, square_root, average, calculate_expression
+        
+        print("  📊 Testing basic arithmetic...")
+        
+        # Test basic operations
+        assert add(5, 3) == 8, "Addition failed"
+        assert subtract(10, 4) == 6, "Subtraction failed"
+        assert multiply(6, 7) == 42, "Multiplication failed"
+        assert divide(15, 3) == 5, "Division failed"
+        print("    ✅ Basic arithmetic: PASSED")
+        
+        # Test advanced operations
+        print("  🧮 Testing advanced operations...")
+        assert factorial(5) == 120, "Factorial failed"
+        assert square_root(16) == 4.0, "Square root failed"
+        assert average([1, 2, 3, 4, 5]) == 3.0, "Average failed"
+        print("    ✅ Advanced operations: PASSED")
+        
+        # Test error handling
+        print("  🚨 Testing error handling...")
+        assert "Error" in str(divide(5, 0)), "Division by zero should return error"
+        assert "Error" in str(factorial(-1)), "Negative factorial should return error"
+        assert "Error" in str(square_root(-1)), "Negative square root should return error"
+        print("    ✅ Error handling: PASSED")
+        
+        # Test expression calculator
+        print("  🧠 Testing expression calculator...")
+        assert calculate_expression("2 + 3 * 4") == 14, "Expression calculation failed"
+        assert calculate_expression("2^3") == 8, "Power expression failed"
+        print("    ✅ Expression calculator: PASSED")
+        
+        # Test quadratic solver
+        print("  📐 Testing quadratic solver...")
+        solutions = MathTools.solve_quadratic(1, -5, 6)
+        assert isinstance(solutions, tuple) and len(solutions) == 2, "Quadratic solver failed"
+        print("    ✅ Quadratic solver: PASSED")
+        
+        # Test practical examples
+        print("  💰 Testing practical calculations...")
+        compound_interest = MathTools.calculate_compound_interest(1000, 0.05, 2)
+        assert compound_interest > 1000, "Compound interest should be greater than principal"
+        
+        percentage_result = MathTools.percentage(25, 100)
+        assert percentage_result == 25.0, "Percentage calculation failed"
+        print("    ✅ Practical calculations: PASSED")
+        
+        print("  ✅ ALL MATH TESTS PASSED!")
+        
+        # Demo output
+        print("\n  📋 Math Tools Demo:")
+        print(f"    5 + 3 = {add(5, 3)}")
+        print(f"    10! = {factorial(10)}")
+        print(f"    √144 = {square_root(144)}")
+        print(f"    Average of [10,20,30] = {average([10,20,30])}")
+        print(f"    Expression '2^3 + 4*5' = {calculate_expression('2^3 + 4*5')}")
+        
+    except ImportError as e:
+        print(f"    ❌ Import error: {e}")
+    except Exception as e:
+        print(f"    ❌ Test failed: {e}")
+        raise
+
+if __name__ == "__main__":
+    test_math_tools()

test_multimodal_tools.py

@@ -0,0 +1,145 @@
+# test_multimodal_tools.py
+import sys
+import os
+from dotenv import load_dotenv
+import tempfile
+import base64
+
+# Load environment variables
+load_dotenv()
+
+# Add tools to path
+sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
+
+def create_test_image():
+    """Create a simple test image for testing"""
+    try:
+        from PIL import Image, ImageDraw
+        
+        # Create a simple test image
+        img = Image.new('RGB', (200, 100), color='white')
+        draw = ImageDraw.Draw(img)
+        draw.text((10, 40), "Test Image", fill='black')
+        
+        # Save to temporary file
+        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.png')
+        img.save(temp_file.name)
+        return temp_file.name
+    except ImportError:
+        print("    ⚠️  PIL not available, creating mock image file")
+        # Create a mock image file for testing
+        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix='.txt')
+        temp_file.write(b"This is a mock image file for testing")
+        temp_file.close()
+        return temp_file.name
+
+def test_multimodal_tools():
+    """Test multimodal tool functions"""
+    try:
+        from tools.multimodal_tools import MultimodalTools, analyze_image, extract_text, analyze_transcript
+        
+        # Check OpenRouter API key
+        openrouter_key = os.getenv("OPENROUTER_API_KEY")
+        if not openrouter_key:
+            print("    ❌ OpenRouter API key not found!")
+            print("    Please add OPENROUTER_API_KEY to your .env file")
+            return
+        
+        print("  🎨 Initializing multimodal tools...")
+        multimodal_tools = MultimodalTools()
+        print(f"    ✅ API key loaded: {openrouter_key[:8]}...")
+        
+        # Test with transcript (doesn't require image)
+        print("  🎙️  Testing audio transcript analysis...")
+        test_transcript = "Hello, this is a test transcript. We're discussing artificial intelligence and machine learning technologies."
+        
+        transcript_result = multimodal_tools.analyze_audio_transcript(
+            test_transcript, 
+            "What is the main topic of this transcript?"
+        )
+        
+        if transcript_result and "Error" not in transcript_result:
+            print("    ✅ Transcript analysis: PASSED")
+            print(f"    Result: {transcript_result[:100]}...")
+        else:
+            print(f"    ❌ Transcript analysis failed: {transcript_result}")
+        
+        # Test convenience function
+        print("  🔧 Testing convenience functions...")
+        convenience_result = analyze_transcript(
+            "This is a test about Python programming and software development.",
+            "Summarize the main topics"
+        )
+        
+        if convenience_result and "Error" not in convenience_result:
+            print("    ✅ Convenience function: PASSED")
+        else:
+            print(f"    ⚠️  Convenience function result: {convenience_result}")
+        
+        # Test image analysis (if we can create a test image)
+        print("  🖼️  Testing image analysis...")
+        try:
+            test_image_path = create_test_image()
+            
+            # Test image description
+            image_result = multimodal_tools.analyze_image(
+                test_image_path,
+                "Describe what you see in this image"
+            )
+            
+            if image_result and "Error" not in image_result:
+                print("    ✅ Image analysis: PASSED")
+                print(f"    Result: {image_result[:100]}...")
+            else:
+                print(f"    ⚠️  Image analysis result: {image_result}")
+            
+            # Test OCR functionality
+            ocr_result = multimodal_tools.extract_text_from_image(test_image_path)
+            if ocr_result and "Error" not in ocr_result:
+                print("    ✅ OCR extraction: PASSED")
+                print(f"    Extracted: {ocr_result[:50]}...")
+            else:
+                print(f"    ⚠️  OCR result: {ocr_result}")
+            
+            # Clean up test image
+            os.unlink(test_image_path)
+            
+        except Exception as e:
+            print(f"    ⚠️  Image testing skipped: {str(e)}")
+        
+        # Test error handling
+        print("  🚨 Testing error handling...")
+        error_result = multimodal_tools.analyze_image(
+            "nonexistent_file.jpg",
+            "This should fail"
+        )
+        
+        if "Error" in error_result:
+            print("    ✅ Error handling: PASSED")
+        else:
+            print("    ⚠️  Error handling unexpected result")
+        
+        # Test empty transcript
+        empty_result = multimodal_tools.analyze_audio_transcript("", "Question")
+        if "Error" in empty_result:
+            print("    ✅ Empty input handling: PASSED")
+        
+        print("  ✅ MULTIMODAL TESTS COMPLETED!")
+        
+        # Demo output
+        print("\n  ���� Multimodal Tools Demo:")
+        demo_transcript = "We discussed the implementation of AI agents using CrewAI framework for automated workflows."
+        demo_result = analyze_transcript(demo_transcript, "What framework was mentioned?")
+        print(f"    Question: What framework was mentioned?")
+        print(f"    Answer: {demo_result[:150]}...")
+        
+    except ImportError as e:
+        print(f"    ❌ Import error: {e}")
+        print("    Make sure all dependencies are installed")
+    except Exception as e:
+        print(f"    ❌ Test failed: {e}")
+        import traceback
+        traceback.print_exc()
+
+if __name__ == "__main__":
+    test_multimodal_tools()

test_search_tools.py

@@ -0,0 +1,96 @@
+# test_search_tools.py
+import sys
+import os
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+# Add tools to path
+sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
+
+def test_search_tools():
+    """Test search tool functions"""
+    try:
+        from tools.search_tools import SearchTools, search_web, search_news
+        
+        print("  🔍 Initializing search tools...")
+        search_tools = SearchTools()
+        
+        # Test DuckDuckGo search (free)
+        print("  🦆 Testing DuckDuckGo search...")
+        ddg_results = search_tools.search_duckduckgo("Python programming", max_results=3)
+        if ddg_results:
+            print(f"    ✅ DuckDuckGo returned {len(ddg_results)} results")
+            for i, result in enumerate(ddg_results[:2]):
+                print(f"      {i+1}. {result.get('title', 'No title')[:50]}...")
+        else:
+            print("    ⚠️  DuckDuckGo returned no results")
+        
+        # Test Tavily search (if API key available)
+        tavily_key = os.getenv("TAVILY_API_KEY")
+        if tavily_key:
+            print("  🔎 Testing Tavily search...")
+            tavily_results = search_tools.search_tavily("AI news 2025", max_results=3)
+            if tavily_results:
+                print(f"    ✅ Tavily returned {len(tavily_results)} results")
+            else:
+                print("    ⚠️  Tavily returned no results")
+        else:
+            print("  ⚠️  Tavily API key not found, skipping Tavily tests")
+        
+        # Test SerpAPI search (if API key available)
+        serpapi_key = os.getenv("SERPAPI_KEY")
+        if serpapi_key:
+            print("  🐍 Testing SerpAPI search...")
+            serp_results = search_tools.search_serpapi("machine learning", max_results=2)
+            if serp_results:
+                print(f"    ✅ SerpAPI returned {len(serp_results)} results")
+            else:
+                print("    ⚠️  SerpAPI returned no results")
+        else:
+            print("  ⚠️  SerpAPI key not found, skipping SerpAPI tests")
+        
+        # Test main search function (with fallback)
+        print("  🎯 Testing main search function...")
+        main_results = search_tools.search("CrewAI framework", max_results=3)
+        if main_results:
+            print(f"    ✅ Main search returned {len(main_results)} results")
+            print(f"    Provider used: {main_results[0].get('source', 'Unknown')}")
+        else:
+            print("    ❌ Main search failed")
+        
+        # Test convenience functions
+        print("  🚀 Testing convenience functions...")
+        web_results = search_web("OpenRouter API", max_results=2)
+        news_results = search_news("artificial intelligence", max_results=2)
+        
+        if web_results:
+            print(f"    ✅ Web search returned {len(web_results)} results")
+        if news_results:
+            print(f"    ✅ News search returned {len(news_results)} results")
+        
+        # Test specialized searches
+        print("  🎓 Testing specialized searches...")
+        academic_results = search_tools.search_academic("machine learning algorithms", max_results=2)
+        if academic_results:
+            print(f"    ✅ Academic search returned {len(academic_results)} results")
+        
+        print("  ✅ ALL SEARCH TESTS COMPLETED!")
+        
+        # Demo output
+        if main_results:
+            print("\n  📋 Search Results Demo:")
+            for i, result in enumerate(main_results[:2]):
+                title = result.get('title', 'No title')[:60]
+                source = result.get('source', 'Unknown')
+                print(f"    {i+1}. [{source}] {title}...")
+        
+    except ImportError as e:
+        print(f"    ❌ Import error: {e}")
+    except Exception as e:
+        print(f"    ❌ Test failed: {e}")
+        print(f"    Error details: {str(e)}")
+
+if __name__ == "__main__":
+    test_search_tools()

test_youtube_tools.py

@@ -0,0 +1,111 @@
+# test_youtube_tools.py (Updated with working test URLs)
+import sys
+import os
+import tempfile
+import shutil
+from dotenv import load_dotenv
+
+load_dotenv()
+sys.path.append(os.path.join(os.path.dirname(__file__), '..'))
+
+def test_youtube_tools():
+    """Test YouTube tool functions with reliable test URLs"""
+    
+    # Updated test URLs - using more reliable, public content
+    test_video_url = "https://www.youtube.com/watch?v=dQw4w9WgXcQ"  # Rick Roll
+    # Use a known public playlist
+    test_playlist_url = "https://www.youtube.com/playlist?list=PLNVhQQnz1F5f11NXuOKF0oULKWWM2HfZS"  # Small public playlist
+    
+    try:
+        from tools.youtube_tools import YouTubeTools, get_video_info, download_video, get_captions
+        
+        print("  📺 Initializing YouTube tools...")
+        youtube_tools = YouTubeTools(max_retries=2, retry_delay=1.0)  # Configure retry settings
+        
+        # Test video info retrieval
+        print("  📋 Testing video info retrieval...")
+        video_info = youtube_tools.get_video_info(test_video_url)
+        
+        if video_info:
+            print("    ✅ Video info retrieval: PASSED")
+            print(f"    Title: {video_info.get('title', 'Unknown')[:50]}...")
+            print(f"    Author: {video_info.get('author', 'Unknown')}")
+            print(f"    Duration: {video_info.get('length', 0)} seconds")
+            print(f"    Views: {video_info.get('views', 0):,}")
+            print(f"    Available captions: {video_info.get('captions_available', [])}")
+        else:
+            print("    ❌ Video info retrieval failed")
+        
+        # Test available qualities with retry
+        print("  🎥 Testing available qualities...")
+        qualities = youtube_tools.get_available_qualities(test_video_url)
+        
+        if qualities:
+            print("    ✅ Quality detection: PASSED")
+            resolutions = [q['resolution'] for q in qualities if q['resolution'] != 'unknown'][:3]
+            print(f"    Available resolutions: {resolutions}")
+        else:
+            print("    ⚠️  Quality detection failed (network issue)")
+        
+        # Test captions (should work without deprecation warning)
+        print("  📝 Testing caption retrieval...")
+        captions = youtube_tools.get_captions(test_video_url, 'en')
+        
+        if captions:
+            print("    ✅ Caption retrieval: PASSED")
+            print(f"    Caption preview: {captions[:100]}...")
+        elif captions is None:
+            print("    ⚠️  No captions available for this video")
+        else:
+            print("    ❌ Caption retrieval failed")
+        
+        # Test playlist info with better URL
+        print("  📂 Testing playlist info...")
+        try:
+            playlist_info = youtube_tools.get_playlist_info(test_playlist_url)
+            
+            if playlist_info:
+                print("    ✅ Playlist info: PASSED")
+                print(f"    Playlist: {playlist_info.get('title', 'Unknown')[:50]}...")
+                print(f"    Videos: {playlist_info.get('video_count', 0)}")
+            else:
+                print("    ⚠️  Playlist info failed")
+        except Exception as e:
+            print(f"    ⚠️  Playlist test error: {str(e)}")
+        
+        # Test convenience functions
+        print("  🔧 Testing convenience functions...")
+        convenience_info = get_video_info(test_video_url)
+        if convenience_info:
+            print("    ✅ Convenience functions: PASSED")
+        
+        # Test error handling
+        print("  🚨 Testing error handling...")
+        invalid_info = youtube_tools.get_video_info("https://invalid-url")
+        if invalid_info is None:
+            print("    ✅ Error handling: PASSED")
+        
+        print("  ✅ YOUTUBE TOOLS TESTS COMPLETED!")
+        
+        # Demo output
+        if video_info:
+            print("\n  📋 YouTube Tools Demo:")
+            print(f"    Video: {video_info.get('title', 'Unknown')}")
+            print(f"    Channel: {video_info.get('author', 'Unknown')}")
+            print(f"    Published: {video_info.get('publish_date', 'Unknown')}")
+            print(f"    Length: {video_info.get('length', 0)} seconds")
+            
+            if qualities:
+                best_quality = qualities[0] if qualities else {}
+                print(f"    Best quality: {best_quality.get('resolution', 'unknown')}")
+        
+    except ImportError as e:
+        print(f"    ❌ Import error: {e}")
+        print("    Make sure pytubefix is installed: pip install pytubefix")
+    except Exception as e:
+        print(f"    ❌ Test failed: {e}")
+        import traceback
+        traceback.print_exc()
+
+if __name__ == "__main__":
+    test_youtube_tools()

tools/__init__.py

@@ -0,0 +1,46 @@
+# tools/__init__.py
+"""
+Tools package for AI agents
+Provides multimodal, search, math, and YouTube capabilities
+"""
+
+from .multimodal_tools import MultimodalTools, analyze_image, extract_text, analyze_transcript
+from .search_tools import SearchTools, search_web, search_news
+from .math_tools import MathTools, add, subtract, multiply, divide, power, factorial, square_root, percentage, average, calculate_expression
+from .youtube_tools import YouTubeTools, get_video_info, download_video, download_audio, get_captions, get_playlist_info
+
+__all__ = [
+    # Multimodal tools
+    'MultimodalTools',
+    'analyze_image',
+    'extract_text', 
+    'analyze_transcript',
+    
+    # Search tools
+    'SearchTools',
+    'search_web',
+    'search_news',
+    
+    # Math tools
+    'MathTools',
+    'add',
+    'subtract', 
+    'multiply',
+    'divide',
+    'power',
+    'factorial',
+    'square_root',
+    'percentage',
+    'average',
+    'calculate_expression',
+    
+    # YouTube tools
+    'YouTubeTools',
+    'get_video_info',
+    'download_video',
+    'download_audio',
+    'get_captions',
+    'get_playlist_info'
+]
+
+__version__ = "1.0.0"

tools/langchain_tools.py

@@ -0,0 +1,128 @@
+# tools/langchain_tools.py (Updated)
+"""
+LangChain-compatible tool wrappers for our existing tools
+"""
+
+from langchain_core.tools import tool
+from typing import Optional
+import os
+from dotenv import load_dotenv
+
+# Load environment variables FIRST, before any tool imports
+load_dotenv()
+
+from .multimodal_tools import MultimodalTools, analyze_transcript as _analyze_transcript
+from .search_tools import SearchTools
+from .math_tools import MathTools
+from .youtube_tools import YouTubeTools
+
+# Initialize tool instances (now env vars are available)
+multimodal_tools = MultimodalTools()
+search_tools = SearchTools()
+youtube_tools = YouTubeTools()
+
+# Rest of the file remains the same...
+@tool
+def extract_text(image_path: str) -> str:
+    """Extract text from an image using OCR"""
+    return multimodal_tools.extract_text_from_image(image_path)
+
+@tool
+def analyze_image_tool(image_path: str, question: str = "Describe this image in detail") -> str:
+    """Analyze an image and answer questions about it"""
+    return multimodal_tools.analyze_image(image_path, question)
+
+@tool
+def analyze_audio_tool(transcript: str, question: str = "Summarize this audio content") -> str:
+    """Analyze audio content via transcript"""
+    return multimodal_tools.analyze_audio_transcript(transcript, question)
+
+@tool
+def search_tool(query: str, max_results: int = 5) -> str:
+    """Search the web for information"""
+    results = search_tools.search(query, max_results)
+    if not results:
+        return "No search results found"
+    
+    # Format results for the LLM
+    formatted_results = []
+    for i, result in enumerate(results, 1):
+        title = result.get('title', 'No title')
+        content = result.get('content', 'No content')
+        url = result.get('url', 'No URL')
+        formatted_results.append(f"{i}. {title}\n{content[:200]}...\nSource: {url}\n")
+    
+    return "\n".join(formatted_results)
+
+@tool
+def extract_youtube_transcript(url: str, language_code: str = 'en') -> str:
+    """Extract transcript/captions from a YouTube video"""
+    captions = youtube_tools.get_captions(url, language_code)
+    if captions:
+        return captions
+    else:
+        return "No captions available for this video"
+
+@tool
+def add(a: float, b: float) -> float:
+    """Add two numbers"""
+    return MathTools.add(a, b)
+
+@tool
+def subtract(a: float, b: float) -> float:
+    """Subtract two numbers"""
+    return MathTools.subtract(a, b)
+
+@tool
+def multiply(a: float, b: float) -> float:
+    """Multiply two numbers"""
+    return MathTools.multiply(a, b)
+
+@tool
+def divide(a: float, b: float) -> str:
+    """Divide two numbers"""
+    result = MathTools.divide(a, b)
+    return str(result)
+
+@tool
+def get_youtube_info(url: str) -> str:
+    """Get information about a YouTube video"""
+    info = youtube_tools.get_video_info(url)
+    if info:
+        return f"Title: {info.get('title', 'Unknown')}\nAuthor: {info.get('author', 'Unknown')}\nDuration: {info.get('length', 0)} seconds\nViews: {info.get('views', 0):,}"
+    else:
+        return "Could not retrieve video information"
+
+@tool
+def calculate_expression(expression: str) -> str:
+    """Calculate a mathematical expression safely"""
+    from .math_tools import calculate_expression as calc_expr
+    return str(calc_expr(expression))
+
+@tool
+def factorial(n: int) -> str:
+    """Calculate factorial of a number"""
+    result = MathTools.factorial(n)
+    return str(result)
+
+@tool
+def square_root(n: float) -> str:
+    """Calculate square root of a number"""
+    result = MathTools.square_root(n)
+    return str(result)
+
+@tool
+def percentage(part: float, whole: float) -> str:
+    """Calculate percentage"""
+    result = MathTools.percentage(part, whole)
+    return str(result)
+
+@tool
+def average(numbers: str) -> str:
+    """Calculate average of numbers (provide as comma-separated string)"""
+    try:
+        number_list = [float(x.strip()) for x in numbers.split(',')]
+        result = MathTools.average(number_list)
+        return str(result)
+    except Exception as e:
+        return f"Error parsing numbers: {str(e)}"

tools/math_tools.py

@@ -0,0 +1,206 @@
+# tools/math_tools.py
+import math
+from typing import Union, List, Any
+from .utils import logger
+
+Number = Union[int, float]
+
+class MathTools:
+    """Simple math tools for basic calculations and utilities"""
+    
+    @staticmethod
+    def add(a: Number, b: Number) -> Number:
+        """Return the sum of a and b"""
+        return a + b
+    
+    @staticmethod
+    def subtract(a: Number, b: Number) -> Number:
+        """Return the difference of a and b"""
+        return a - b
+    
+    @staticmethod
+    def multiply(a: Number, b: Number) -> Number:
+        """Return the product of a and b"""
+        return a * b
+    
+    @staticmethod
+    def divide(a: Number, b: Number) -> Union[Number, str]:
+        """Return the division of a by b, handle division by zero"""
+        if b == 0:
+            return 'Error: Division by zero'
+        return a / b
+    
+    @staticmethod
+    def power(base: Number, exponent: Number) -> Number:
+        """Return base raised to the power of exponent"""
+        return base ** exponent
+    
+    @staticmethod
+    def factorial(n: int) -> Union[int, str]:
+        """Return factorial of n (non-negative integer)"""
+        if not isinstance(n, int) or n < 0:
+            return 'Error: Input must be a non-negative integer'
+        if n == 0 or n == 1:
+            return 1
+        result = 1
+        for i in range(2, n + 1):
+            result *= i
+        return result
+    
+    @staticmethod
+    def square_root(n: Number) -> Union[float, str]:
+        """Return square root of n"""
+        if n < 0:
+            return 'Error: Cannot calculate square root of negative number'
+        return math.sqrt(n)
+    
+    @staticmethod
+    def percentage(part: Number, whole: Number) -> Union[float, str]:
+        """Calculate percentage: (part/whole) * 100"""
+        if whole == 0:
+            return 'Error: Cannot calculate percentage with zero denominator'
+        return (part / whole) * 100
+    
+    @staticmethod
+    def average(numbers: List[Number]) -> Union[float, str]:
+        """Calculate average of a list of numbers"""
+        if not numbers:
+            return 'Error: Cannot calculate average of empty list'
+        return sum(numbers) / len(numbers)
+    
+    @staticmethod
+    def round_number(n: Number, decimals: int = 2) -> Number:
+        """Round number to specified decimal places"""
+        return round(n, decimals)
+    
+    @staticmethod
+    def absolute(n: Number) -> Number:
+        """Return absolute value of n"""
+        return abs(n)
+    
+    @staticmethod
+    def min_value(numbers: List[Number]) -> Union[Number, str]:
+        """Find minimum value in list"""
+        if not numbers:
+            return 'Error: Cannot find minimum of empty list'
+        return min(numbers)
+    
+    @staticmethod
+    def max_value(numbers: List[Number]) -> Union[Number, str]:
+        """Find maximum value in list"""
+        if not numbers:
+            return 'Error: Cannot find maximum of empty list'
+        return max(numbers)
+    
+    @staticmethod
+    def calculate_compound_interest(principal: Number, rate: Number, time: Number, compounds_per_year: int = 1) -> float:
+        """
+        Calculate compound interest
+        Formula: A = P(1 + r/n)^(nt)
+        """
+        return principal * (1 + rate/compounds_per_year) ** (compounds_per_year * time)
+    
+    @staticmethod
+    def solve_quadratic(a: Number, b: Number, c: Number) -> Union[tuple, str]:
+        """
+        Solve quadratic equation ax² + bx + c = 0
+        Returns tuple of solutions or error message
+        """
+        if a == 0:
+            return 'Error: Not a quadratic equation (a cannot be 0)'
+        
+        discriminant = b**2 - 4*a*c
+        
+        if discriminant < 0:
+            return 'Error: No real solutions (negative discriminant)'
+        elif discriminant == 0:
+            solution = -b / (2*a)
+            return (solution, solution)
+        else:
+            sqrt_discriminant = math.sqrt(discriminant)
+            solution1 = (-b + sqrt_discriminant) / (2*a)
+            solution2 = (-b - sqrt_discriminant) / (2*a)
+            return (solution1, solution2)
+
+# Convenience functions for direct use
+def add(a: Number, b: Number) -> Number:
+    """Add two numbers"""
+    return MathTools.add(a, b)
+
+def subtract(a: Number, b: Number) -> Number:
+    """Subtract two numbers"""
+    return MathTools.subtract(a, b)
+
+def multiply(a: Number, b: Number) -> Number:
+    """Multiply two numbers"""
+    return MathTools.multiply(a, b)
+
+def divide(a: Number, b: Number) -> Union[Number, str]:
+    """Divide two numbers"""
+    return MathTools.divide(a, b)
+
+def power(base: Number, exponent: Number) -> Number:
+    """Raise base to power of exponent"""
+    return MathTools.power(base, exponent)
+
+def factorial(n: int) -> Union[int, str]:
+    """Calculate factorial of n"""
+    return MathTools.factorial(n)
+
+def square_root(n: Number) -> Union[float, str]:
+    """Calculate square root"""
+    return MathTools.square_root(n)
+
+def percentage(part: Number, whole: Number) -> Union[float, str]:
+    """Calculate percentage"""
+    return MathTools.percentage(part, whole)
+
+def average(numbers: List[Number]) -> Union[float, str]:
+    """Calculate average of numbers"""
+    return MathTools.average(numbers)
+
+def calculate_expression(expression: str) -> Union[Number, str]:
+    """
+    Safely evaluate mathematical expressions
+    WARNING: Only use with trusted input
+    """
+    try:
+        # Only allow safe mathematical operations
+        allowed_chars = set('0123456789+-*/().^ ')
+        if not all(c in allowed_chars for c in expression.replace('**', '^')):
+            return 'Error: Invalid characters in expression'
+        
+        # Replace ^ with ** for Python exponentiation
+        safe_expression = expression.replace('^', '**')
+        
+        # Evaluate the expression
+        result = eval(safe_expression)
+        return result
+    except ZeroDivisionError:
+        return 'Error: Division by zero in expression'
+    except Exception as e:
+        return f'Error: Invalid expression - {str(e)}'
+
+# Example usage and testing
+if __name__ == "__main__":
+    # Test basic operations
+    print("Basic Operations:")
+    print(f"5 + 3 = {add(5, 3)}")
+    print(f"10 - 4 = {subtract(10, 4)}")
+    print(f"6 * 7 = {multiply(6, 7)}")
+    print(f"15 / 3 = {divide(15, 3)}")
+    print(f"2^8 = {power(2, 8)}")
+    
+    print("\nAdvanced Operations:")
+    print(f"√16 = {square_root(16)}")
+    print(f"5! = {factorial(5)}")
+    print(f"Average of [1,2,3,4,5] = {average([1,2,3,4,5])}")
+    percent_result = percentage(25, 100)
+    if isinstance(percent_result, float):
+        print(f"25% of 200 = {percent_result * 200 / 100}")
+    else:
+        print(f"25% of 200 = {percent_result}")
+    
+    print("\nQuadratic Equation (x² - 5x + 6 = 0):")
+    solutions = MathTools.solve_quadratic(1, -5, 6)
+    print(f"Solutions: {solutions}")

tools/multimodal_tools.py

@@ -0,0 +1,166 @@
+# tools/multimodal_tools.py
+import requests
+import json
+from typing import Optional, Dict, Any
+from .utils import encode_image_to_base64, validate_file_exists, get_env_var, logger
+
+class MultimodalTools:
+    """Free multimodal AI tools using OpenRouter and other free services"""
+    
+    def __init__(self, openrouter_key: Optional[str] = None):
+        self.openrouter_key = openrouter_key or get_env_var("OPENROUTER_API_KEY", None)
+        self.openrouter_url = "https://openrouter.ai/api/v1/chat/completions"
+        self.headers = {
+            "Authorization": f"Bearer {self.openrouter_key}",
+            "Content-Type": "application/json",
+            "HTTP-Referer": "https://your-app.com",  # Optional: for analytics
+            "X-Title": "Multimodal Tools"  # Optional: for analytics
+        }
+        
+        # Available free multimodal models
+        self.vision_model = "moonshotai/kimi-vl-a3b-thinking:free"
+        self.text_model = "meta-llama/llama-4-maverick:free"
+    
+    def _make_openrouter_request(self, payload: Dict[str, Any]) -> str:
+        """Make request to OpenRouter API with error handling"""
+        try:
+            response = requests.post(
+                self.openrouter_url, 
+                headers=self.headers, 
+                json=payload,
+                timeout=30
+            )
+            response.raise_for_status()
+            
+            result = response.json()
+            if 'choices' in result and len(result['choices']) > 0:
+                return result['choices'][0]['message']['content']
+            else:
+                logger.error(f"Unexpected response format: {result}")
+                return "Error: Invalid response format"
+                
+        except requests.exceptions.RequestException as e:
+            logger.error(f"OpenRouter API request failed: {str(e)}")
+            return f"Error making API request: {str(e)}"
+        except Exception as e:
+            logger.error(f"Unexpected error: {str(e)}")
+            return f"Unexpected error: {str(e)}"
+    
+    def analyze_image(self, image_path: str, question: str = "Describe this image in detail") -> str:
+        """
+        Analyze image content using multimodal AI
+        
+        Args:
+            image_path: Path to image file
+            question: Question about the image
+            
+        Returns:
+            AI analysis of the image
+        """
+        if not validate_file_exists(image_path):
+            return f"Error: Image file not found at {image_path}"
+        
+        try:
+            encoded_image = encode_image_to_base64(image_path)
+            
+            payload = {
+                "model": self.vision_model,
+                "messages": [
+                    {
+                        "role": "user", 
+                        "content": [
+                            {"type": "text", "text": question},
+                            {
+                                "type": "image_url", 
+                                "image_url": {"url": f"data:image/jpeg;base64,{encoded_image}"}
+                            }
+                        ]
+                    }
+                ],
+                "temperature": 0,
+                "max_tokens": 1024
+            }
+            
+            return self._make_openrouter_request(payload)
+            
+        except Exception as e:
+            error_msg = f"Error analyzing image: {str(e)}"
+            logger.error(error_msg)
+            return error_msg
+    
+    def extract_text_from_image(self, image_path: str) -> str:
+        """
+        Extract text from image using OCR via multimodal AI
+        
+        Args:
+            image_path: Path to image file
+            
+        Returns:
+            Extracted text from image
+        """
+        ocr_prompt = """Extract all visible text from this image. 
+        Return only the text content without any additional commentary or formatting. 
+        If no text is visible, return 'No text found'."""
+        
+        return self.analyze_image(image_path, ocr_prompt)
+    
+    def analyze_audio_transcript(self, transcript: str, question: str = "Summarize this audio content") -> str:
+        """
+        Analyze audio content via transcript
+        
+        Args:
+            transcript: Audio transcript text
+            question: Question about the audio content
+            
+        Returns:
+            AI analysis of the audio content
+        """
+        if not transcript.strip():
+            return "Error: Empty transcript provided"
+        
+        try:
+            payload = {
+                "model": self.text_model,
+                "messages": [
+                    {
+                        "role": "user", 
+                        "content": f"Audio transcript: {transcript}\n\nQuestion: {question}"
+                    }
+                ],
+                "temperature": 0,
+                "max_tokens": 1024
+            }
+            
+            return self._make_openrouter_request(payload)
+            
+        except Exception as e:
+            error_msg = f"Error analyzing audio transcript: {str(e)}"
+            logger.error(error_msg)
+            return error_msg
+    
+    def describe_image(self, image_path: str) -> str:
+        """Get a detailed description of an image"""
+        return self.analyze_image(
+            image_path, 
+            "Provide a detailed, objective description of this image including objects, people, colors, setting, and any notable details."
+        )
+    
+    def answer_visual_question(self, image_path: str, question: str) -> str:
+        """Answer a specific question about an image"""
+        return self.analyze_image(image_path, question)
+
+# Convenience functions for direct use
+def analyze_image(image_path: str, question: str = "Describe this image in detail") -> str:
+    """Standalone function to analyze an image"""
+    tools = MultimodalTools()
+    return tools.analyze_image(image_path, question)
+
+def extract_text(image_path: str) -> str:
+    """Standalone function to extract text from an image"""
+    tools = MultimodalTools()
+    return tools.extract_text_from_image(image_path)
+
+def analyze_transcript(transcript: str, question: str = "Summarize this content") -> str:
+    """Standalone function to analyze audio transcript"""
+    tools = MultimodalTools()
+    return tools.analyze_audio_transcript(transcript, question)

tools/search_tools.py

@@ -0,0 +1,223 @@
+# tools/search_tools.py
+import requests
+import os
+from typing import List, Dict, Any, Optional
+from .utils import get_env_var, logger
+
+class SearchTools:
+    """Free and cost-effective search tools with multiple providers"""
+    
+    def __init__(self):
+        # Primary: Free alternatives
+        self.duckduckgo_enabled = True
+        
+        # Secondary: Tavily (cost-effective)
+        self.tavily_api_key = os.getenv("TAVILY_API_KEY")
+        
+        # Tertiary: SerpAPI (expensive, fallback only)
+        self.serpapi_key = os.getenv("SERPAPI_KEY")
+    
+    def search_duckduckgo(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+        """
+        Free search using DuckDuckGo Instant Answer API
+        
+        Args:
+            query: Search query
+            max_results: Maximum number of results
+            
+        Returns:
+            List of search results
+        """
+        try:
+            # DuckDuckGo Instant Answer API (free)
+            url = "https://api.duckduckgo.com/"
+            params = {
+                'q': query,
+                'format': 'json',
+                'no_html': '1',
+                'skip_disambig': '1'
+            }
+            
+            response = requests.get(url, params=params, timeout=10)
+            response.raise_for_status()
+            
+            data = response.json()
+            results = []
+            
+            # Process abstract
+            if data.get('Abstract'):
+                results.append({
+                    'title': data.get('Heading', 'DuckDuckGo Result'),
+                    'url': data.get('AbstractURL', ''),
+                    'content': data.get('Abstract', ''),
+                    'source': 'DuckDuckGo'
+                })
+            
+            # Process related topics
+            for topic in data.get('RelatedTopics', [])[:max_results-len(results)]:
+                if isinstance(topic, dict) and 'Text' in topic:
+                    results.append({
+                        'title': topic.get('Text', '')[:100],
+                        'url': topic.get('FirstURL', ''),
+                        'content': topic.get('Text', ''),
+                        'source': 'DuckDuckGo'
+                    })
+            
+            return results[:max_results]
+            
+        except Exception as e:
+            logger.error(f"DuckDuckGo search failed: {str(e)}")
+            return []
+    
+    def search_tavily(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+        """
+        Search using Tavily API (cost-effective)
+        
+        Args:
+            query: Search query
+            max_results: Maximum number of results
+            
+        Returns:
+            List of search results
+        """
+        if not self.tavily_api_key:
+            logger.warning("Tavily API key not provided")
+            return []
+        
+        try:
+            url = "https://api.tavily.com/search"
+            payload = {
+                "api_key": self.tavily_api_key,
+                "query": query,
+                "search_depth": "basic",
+                "include_answer": False,
+                "include_images": False,
+                "include_raw_content": False,
+                "max_results": max_results
+            }
+            
+            response = requests.post(url, json=payload, timeout=15)
+            response.raise_for_status()
+            
+            data = response.json()
+            results = []
+            
+            for result in data.get('results', []):
+                results.append({
+                    'title': result.get('title', ''),
+                    'url': result.get('url', ''),
+                    'content': result.get('content', ''),
+                    'source': 'Tavily'
+                })
+            
+            return results
+            
+        except Exception as e:
+            logger.error(f"Tavily search failed: {str(e)}")
+            return []
+    
+    def search_serpapi(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+        """
+        Search using SerpAPI (expensive, fallback only)
+        
+        Args:
+            query: Search query
+            max_results: Maximum number of results
+            
+        Returns:
+            List of search results
+        """
+        if not self.serpapi_key:
+            logger.warning("SerpAPI key not provided")
+            return []
+        
+        try:
+            url = "https://serpapi.com/search"
+            params = {
+                'api_key': self.serpapi_key,
+                'engine': 'google',
+                'q': query,
+                'num': max_results,
+                'gl': 'us',  # Geolocation
+                'hl': 'en'   # Language
+            }
+            
+            response = requests.get(url, params=params, timeout=15)
+            response.raise_for_status()
+            
+            data = response.json()
+            results = []
+            
+            for result in data.get('organic_results', []):
+                results.append({
+                    'title': result.get('title', ''),
+                    'url': result.get('link', ''),
+                    'content': result.get('snippet', ''),
+                    'source': 'Google (SerpAPI)'
+                })
+            
+            return results
+            
+        except Exception as e:
+            logger.error(f"SerpAPI search failed: {str(e)}")
+            return []
+    
+    def search(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+        """
+        Comprehensive search using multiple providers with fallback strategy
+        
+        Args:
+            query: Search query
+            max_results: Maximum number of results
+            
+        Returns:
+            List of search results from best available provider
+        """
+        if not query.strip():
+            return []
+        
+        # Try providers in order of preference (free -> cheap -> expensive)
+        providers = [
+            ("DuckDuckGo", self.search_duckduckgo),
+            ("Tavily", self.search_tavily),
+            ("SerpAPI", self.search_serpapi)
+        ]
+        
+        for provider_name, search_func in providers:
+            try:
+                logger.info(f"Attempting search with {provider_name}")
+                results = search_func(query, max_results)
+                
+                if results:
+                    logger.info(f"Successfully retrieved {len(results)} results from {provider_name}")
+                    return results
+                else:
+                    logger.warning(f"No results from {provider_name}")
+                    
+            except Exception as e:
+                logger.error(f"Error with {provider_name}: {str(e)}")
+                continue
+        
+        logger.error("All search providers failed")
+        return []
+    
+    def search_news(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+        """Search for news articles"""
+        news_query = f"news {query}"
+        return self.search(news_query, max_results)
+    
+    def search_academic(self, query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+        """Search for academic content"""
+        academic_query = f"academic research {query} site:scholar.google.com OR site:arxiv.org OR site:researchgate.net"
+        return self.search(academic_query, max_results)
+
+# Convenience functions
+def search_web(query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+    """Standalone function for web search"""
+    tools = SearchTools()
+    return tools.search(query, max_results)
+
+def search_news(query: str, max_results: int = 5) -> List[Dict[str, Any]]:
+    """Standalone function for news search"""
+    tools = SearchTools()
+    return tools.search_news(query, max_results)

tools/utils.py

@@ -0,0 +1,36 @@
+# tools/utils.py
+import base64
+import os
+import mimetypes
+from typing import Optional, Dict, Any
+import logging
+
+# Configure logging
+logging.basicConfig(level=logging.INFO)
+logger = logging.getLogger(__name__)
+
+def encode_image_to_base64(image_path: str) -> str:
+    """Convert image file to base64 encoding"""
+    try:
+        with open(image_path, "rb") as image_file:
+            encoded = base64.b64encode(image_file.read()).decode('utf-8')
+            return encoded
+    except Exception as e:
+        logger.error(f"Error encoding image {image_path}: {str(e)}")
+        raise
+
+def get_mime_type(file_path: str) -> str:
+    """Get MIME type for file"""
+    mime_type, _ = mimetypes.guess_type(file_path)
+    return mime_type or 'application/octet-stream'
+
+def validate_file_exists(file_path: str) -> bool:
+    """Validate that file exists and is readable"""
+    return os.path.exists(file_path) and os.path.isfile(file_path)
+
+def get_env_var(var_name: str, default: Optional[str] = None) -> str:
+    """Get environment variable with optional default"""
+    value = os.getenv(var_name, default)
+    if value is None:
+        raise ValueError(f"Environment variable {var_name} is required")
+    return value

tools/youtube_tools.py

@@ -0,0 +1,315 @@
+# tools/youtube_tools.py (Updated with fixes)
+"""
+YouTube Tools Module - Fixed version using pytubefix
+Addresses network issues, deprecation warnings, and playlist errors
+"""
+
+from pytubefix import YouTube, Playlist
+from pytubefix.cli import on_progress
+from typing import Optional, Dict, Any, List
+import os
+import time
+import logging
+from .utils import logger, validate_file_exists
+
+class YouTubeTools:
+    """YouTube tools with improved error handling and network resilience"""
+    
+    def __init__(self, max_retries: int = 3, retry_delay: float = 1.0):
+        self.supported_formats = ['mp4', '3gp', 'webm']
+        self.supported_audio_formats = ['mp3', 'mp4', 'webm']
+        self.max_retries = max_retries
+        self.retry_delay = retry_delay
+    
+    def _retry_operation(self, operation, *args, **kwargs):
+        """Retry operation with exponential backoff for network issues"""
+        for attempt in range(self.max_retries):
+            try:
+                return operation(*args, **kwargs)
+            except Exception as e:
+                if attempt == self.max_retries - 1:
+                    raise e
+                
+                error_msg = str(e).lower()
+                if any(term in error_msg for term in ['network', 'socket', 'timeout', 'connection']):
+                    wait_time = self.retry_delay * (2 ** attempt)
+                    logger.warning(f"Network error (attempt {attempt + 1}/{self.max_retries}): {e}")
+                    logger.info(f"Retrying in {wait_time} seconds...")
+                    time.sleep(wait_time)
+                else:
+                    raise e
+    
+    def get_video_info(self, url: str) -> Optional[Dict[str, Any]]:
+        """
+        Retrieve comprehensive metadata about a YouTube video using pytubefix
+        """
+        try:
+            def _get_info():
+                yt = YouTube(url, on_progress_callback=on_progress)
+                
+                # Get available streams info with better error handling
+                video_streams = []
+                try:
+                    streams = yt.streams.filter(progressive=True, file_extension='mp4')
+                    for stream in streams:
+                        try:
+                            video_streams.append({
+                                'resolution': getattr(stream, 'resolution', 'unknown'),
+                                'fps': getattr(stream, 'fps', 'unknown'),
+                                'video_codec': getattr(stream, 'video_codec', 'unknown'),
+                                'audio_codec': getattr(stream, 'audio_codec', 'unknown'),
+                                'filesize': getattr(stream, 'filesize', None),
+                                'mime_type': getattr(stream, 'mime_type', 'unknown')
+                            })
+                        except Exception as stream_error:
+                            logger.debug(f"Error processing stream: {stream_error}")
+                            continue
+                except Exception as e:
+                    logger.warning(f"Could not retrieve stream details: {e}")
+                
+                # Get caption languages safely
+                captions_available = []
+                try:
+                    if yt.captions:
+                        captions_available = list(yt.captions.keys())
+                except Exception as e:
+                    logger.warning(f"Could not retrieve captions list: {e}")
+                
+                info = {
+                    'title': getattr(yt, 'title', 'Unknown'),
+                    'author': getattr(yt, 'author', 'Unknown'),
+                    'channel_url': getattr(yt, 'channel_url', 'Unknown'),
+                    'length': getattr(yt, 'length', 0),
+                    'views': getattr(yt, 'views', 0),
+                    'description': getattr(yt, 'description', ''),
+                    'thumbnail_url': getattr(yt, 'thumbnail_url', ''),
+                    'publish_date': yt.publish_date.isoformat() if getattr(yt, 'publish_date', None) else None,
+                    'keywords': getattr(yt, 'keywords', []),
+                    'video_id': getattr(yt, 'video_id', ''),
+                    'watch_url': getattr(yt, 'watch_url', url),
+                    'available_streams': video_streams,
+                    'captions_available': captions_available
+                }
+                
+                return info
+            
+            info = self._retry_operation(_get_info)
+            if info is not None:
+                logger.info(f"Retrieved info for video: {info.get('title', 'Unknown')}")
+            return info
+            
+        except Exception as e:
+            logger.error(f"Failed to get video info for {url}: {e}")
+            return None
+    
+    def download_video(self, url: str, output_path: str = './downloads', 
+                      resolution: str = 'highest', filename: Optional[str] = None) -> Optional[str]:
+        """Download a YouTube video with retry logic"""
+        try:
+            def _download():
+                os.makedirs(output_path, exist_ok=True)
+                
+                yt = YouTube(url, on_progress_callback=on_progress)
+                
+                # Select stream based on resolution preference
+                if resolution == 'highest':
+                    stream = yt.streams.get_highest_resolution()
+                elif resolution == 'lowest':
+                    stream = yt.streams.get_lowest_resolution()
+                else:
+                    stream = yt.streams.filter(res=resolution, progressive=True, file_extension='mp4').first()
+                    if not stream:
+                        logger.warning(f"Resolution {resolution} not found, downloading highest instead")
+                        stream = yt.streams.get_highest_resolution()
+                
+                if not stream:
+                    raise Exception("No suitable stream found for download")
+                
+                # Download with custom filename if provided
+                if filename:
+                    safe_filename = "".join(c for c in filename if c.isalnum() or c in (' ', '-', '_')).rstrip()
+                    file_path = stream.download(output_path=output_path, filename=f"{safe_filename}.{stream.subtype}")
+                else:
+                    file_path = stream.download(output_path=output_path)
+                
+                return file_path
+            
+            file_path = self._retry_operation(_download)
+            logger.info(f"Downloaded video to {file_path}")
+            return file_path
+            
+        except Exception as e:
+            logger.error(f"Failed to download video from {url}: {e}")
+            return None
+    
+    def download_audio(self, url: str, output_path: str = './downloads', 
+                      filename: Optional[str] = None) -> Optional[str]:
+        """Download only audio from a YouTube video with retry logic"""
+        try:
+            def _download_audio():
+                os.makedirs(output_path, exist_ok=True)
+                
+                yt = YouTube(url, on_progress_callback=on_progress)
+                audio_stream = yt.streams.get_audio_only()
+                
+                if not audio_stream:
+                    raise Exception("No audio stream found")
+                
+                if filename:
+                    safe_filename = "".join(c for c in filename if c.isalnum() or c in (' ', '-', '_')).rstrip()
+                    file_path = audio_stream.download(output_path=output_path, filename=f"{safe_filename}.{audio_stream.subtype}")
+                else:
+                    file_path = audio_stream.download(output_path=output_path)
+                
+                return file_path
+            
+            file_path = self._retry_operation(_download_audio)
+            logger.info(f"Downloaded audio to {file_path}")
+            return file_path
+            
+        except Exception as e:
+            logger.error(f"Failed to download audio from {url}: {e}")
+            return None
+    
+    def get_captions(self, url: str, language_code: str = 'en') -> Optional[str]:
+        """
+        Get captions/subtitles - FIXED: No more deprecation warning
+        """
+        try:
+            def _get_captions():
+                yt = YouTube(url, on_progress_callback=on_progress)
+                
+                if not yt.captions:
+                    logger.warning("No captions available for this video")
+                    return None
+                
+                # Use modern dictionary-style access instead of deprecated method
+                if language_code in yt.captions:
+                    caption = yt.captions[language_code]
+                    captions_text = caption.generate_srt_captions()
+                    return captions_text
+                else:
+                    available_langs = list(yt.captions.keys())
+                    logger.warning(f"Captions not found for language {language_code}. Available: {available_langs}")
+                    return None
+            
+            result = self._retry_operation(_get_captions)
+            if result:
+                logger.info(f"Retrieved captions in {language_code}")
+            return result
+                
+        except Exception as e:
+            logger.error(f"Failed to get captions from {url}: {e}")
+            return None
+    
+    def get_playlist_info(self, playlist_url: str) -> Optional[Dict[str, Any]]:
+        """
+        Get information about a YouTube playlist - FIXED: Better error handling
+        """
+        try:
+            def _get_playlist_info():
+                playlist = Playlist(playlist_url)
+                
+                # Get video URLs first (this triggers the playlist loading)
+                video_urls = list(playlist.video_urls)
+                
+                # Safely access playlist properties with fallbacks
+                info = {
+                    'video_count': len(video_urls),
+                    'video_urls': video_urls[:10],  # Limit to first 10 for performance
+                    'total_videos': len(video_urls)
+                }
+                
+                # Try to get additional info, but don't fail if unavailable
+                try:
+                    info['title'] = getattr(playlist, 'title', 'Unknown Playlist')
+                except:
+                    info['title'] = 'Private/Unavailable Playlist'
+                
+                try:
+                    info['description'] = getattr(playlist, 'description', '')
+                except:
+                    info['description'] = 'Description unavailable'
+                
+                try:
+                    info['owner'] = getattr(playlist, 'owner', 'Unknown')
+                except:
+                    info['owner'] = 'Owner unavailable'
+                
+                return info
+            
+            info = self._retry_operation(_get_playlist_info)
+            if info is not None:
+                logger.info(f"Retrieved playlist info: {info['title']} ({info['video_count']} videos)")
+            return info
+            
+        except Exception as e:
+            logger.error(f"Failed to get playlist info from {playlist_url}: {e}")
+            return None
+    
+    def get_available_qualities(self, url: str) -> Optional[List[Dict[str, Any]]]:
+        """
+        Get all available download qualities - FIXED: Better network handling
+        """
+        try:
+            def _get_qualities():
+                yt = YouTube(url, on_progress_callback=on_progress)
+                streams = []
+                
+                # Get progressive streams (video + audio)
+                for stream in yt.streams.filter(progressive=True):
+                    try:
+                        streams.append({
+                            'resolution': getattr(stream, 'resolution', 'unknown'),
+                            'fps': getattr(stream, 'fps', 'unknown'),
+                            'filesize_mb': round(stream.filesize / (1024 * 1024), 2) if getattr(stream, 'filesize', None) else None,
+                            'mime_type': getattr(stream, 'mime_type', 'unknown'),
+                            'video_codec': getattr(stream, 'video_codec', 'unknown'),
+                            'audio_codec': getattr(stream, 'audio_codec', 'unknown')
+                        })
+                    except Exception as stream_error:
+                        logger.debug(f"Error processing stream: {stream_error}")
+                        continue
+                
+                # Sort by resolution (numeric part)
+                def sort_key(x):
+                    res = x['resolution']
+                    if res and res != 'unknown' and res[:-1].isdigit():
+                        return int(res[:-1])
+                    return 0
+                
+                return sorted(streams, key=sort_key, reverse=True)
+            
+            return self._retry_operation(_get_qualities)
+            
+        except Exception as e:
+            logger.error(f"Failed to get qualities for {url}: {e}")
+            return None
+
+# Convenience functions (unchanged)
+def get_video_info(url: str) -> Optional[Dict[str, Any]]:
+    """Standalone function to get video information"""
+    tools = YouTubeTools()
+    return tools.get_video_info(url)
+
+def download_video(url: str, output_path: str = './downloads', 
+                  resolution: str = 'highest', filename: Optional[str] = None) -> Optional[str]:
+    """Standalone function to download a video"""
+    tools = YouTubeTools()
+    return tools.download_video(url, output_path, resolution, filename)
+
+def download_audio(url: str, output_path: str = './downloads', 
+                  filename: Optional[str] = None) -> Optional[str]:
+    """Standalone function to download audio only"""
+    tools = YouTubeTools()
+    return tools.download_audio(url, output_path, filename)
+
+def get_captions(url: str, language_code: str = 'en') -> Optional[str]:
+    """Standalone function to get video captions"""
+    tools = YouTubeTools()
+    return tools.get_captions(url, language_code)
+
+def get_playlist_info(playlist_url: str) -> Optional[Dict[str, Any]]:
+    """Standalone function to get playlist information"""
+    tools = YouTubeTools()
+    return tools.get_playlist_info(playlist_url)