Update agent.py

agent.py

@@ -28,6 +28,71 @@ openai_api_key = os.getenv("OPENAI_API_KEY")  # Replace with your actual key
 search_tool = DuckDuckGoSearchResults(max_results=3)
 wikipedia = WikipediaQueryRun(api_wrapper=WikipediaAPIWrapper(top_k_results=3))
 
+#video analysis
+@tool
+def analyze_video(video_url: str, question: str) -> str:
+    """
+    Analyze a video to answer a specific question about its content.
+    Supports YouTube videos. Provide both the video URL and the question.
+    
+    Example: 
+    analyze_video(
+        "https://www.youtube.com/watch?v=1htKBjuUWec", 
+        "What does Teal'c say in response to the question 'Isn't that hot?'"
+    )
+    """
+    try:
+        # Create a temporary directory for processing
+        with tempfile.TemporaryDirectory() as tmp_dir:
+            # Download audio from YouTube
+            ydl_opts = {
+                'format': 'bestaudio/best',
+                'outtmpl': os.path.join(tmp_dir, 'audio.%(ext)s'),
+                'postprocessors': [{
+                    'key': 'FFmpegExtractAudio',
+                    'preferredcodec': 'mp3',
+                    'preferredquality': '192',
+                }],
+                'quiet': True,
+            }
+            
+            with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+                ydl.download([video_url])
+                audio_file = os.path.join(tmp_dir, 'audio.mp3')
+                
+                # If the file wasn't created with .mp3 extension, find it
+                if not os.path.exists(audio_file):
+                    for f in os.listdir(tmp_dir):
+                        if f.startswith('audio.'):
+                            audio_file = os.path.join(tmp_dir, f)
+                            break
+            
+            # Transcribe audio using OpenAI Whisper
+            client = OpenAI()
+            with open(audio_file, 'rb') as f:
+                transcript = client.audio.transcriptions.create(
+                    model="whisper-large-v3",
+                    file=f,
+                    response_format="text"
+                )
+            
+            # Use GPT to answer the question based on transcript
+            response = client.chat.completions.create(
+                model="gpt-4-turbo",
+                messages=[
+                    {"role": "system", "content": "Answer the question based on the video transcript below. Be precise and quote directly when possible."},
+                    {"role": "user", "content": f"Question: {question}\n\nTranscript:\n{transcript}"}
+                ],
+                max_tokens=300
+            )
+            
+            return response.choices[0].message.content
+            
+    except Exception as e:
+        return f"Error analyzing video: {str(e)}"
+
+
+
 # Image Description Tool (using GPT-4 Vision)
 @tool
 def describe_image(image_url: str) -> str:
@@ -108,7 +173,7 @@ def convert_currency(amount: float, from_currency: str, to_currency: str) -> flo
 # ---------------------
 
 # Available tools
-tools = [search_tool, wikipedia, describe_image, calculate, convert_date, convert_currency]
+tools = [search_tool, wikipedia, analyze_video, describe_image, calculate, convert_date, convert_currency]
 tools_as_openai = [convert_to_openai_tool(t) for t in tools]
 
 # Agent State Definition