Update langgraph_agent.py

langgraph_agent.py

@@ -5,8 +5,8 @@ import pandas as pd
 from typing import Dict, List, Union
 import re
 
-from PIL import Image as PILImage # Keep PIL for potential future use or if other parts depend on it, but describe_image is removed.
-from huggingface_hub import InferenceClient # Keep InferenceClient for other potential HF uses, but describe_image is removed.
+from PIL import Image as PILImage  # Keep PIL for potential future use or if other parts depend on it, but describe_image is removed.
+from huggingface_hub import InferenceClient  # Keep InferenceClient for other potential HF uses, but describe_image is removed.
 
 from langgraph.graph import START, StateGraph, MessagesState
 from langgraph.prebuilt import tools_condition, ToolNode
@@ -73,8 +73,6 @@ def google_web_search(query: str) -> str:
         return f"Error occurred while searching the web for '{query}'. Details: {str(e)}"
 
 
-# HF_API_TOKEN is no longer directly needed for describe_image as that tool is removed.
-# But keeping InferenceClient initialization for completeness if other HF tools might be added later.
 HF_API_TOKEN = os.getenv("HF_API_TOKEN")
 MODEL = os.getenv("MODEL")
 HF_INFERENCE_CLIENT = None
@@ -133,50 +131,17 @@ def python_interpreter(code: str) -> Dict[str, str]:
         except Exception as e:
             return {"execution_error": str(e)}
 
-# --- Youtube Tool (Remains the same) ---
-@tool
-def Youtube(url: str, question: str) -> Dict[str, str]:
-    """
-    Tells about the YouTube video identified by the given URL, answering a question about it.
-    Note: This is a simulated response. In a real application, this would interact with a YouTube API
-    or a video analysis service to get actual video information and transcripts.
-    """
-    print(f"Youtube called with URL: {url}, Question: {question}")
-    
-    # Placeholder for actual YouTube API call.
-    # In a real scenario, you'd use a library like `google-api-python-client` for YouTube Data API
-    # or a dedicated video transcription/analysis service.
-
-    # Simulating the previous video content for demonstration
-    if "https://www.youtube.com/watch?v=1htKBjuUWec" in url or re.search(r'youtube\.com/watch\?v=|youtu\.be/', url):
-        return {
-            "video_url": url,
-            "question_asked": question,
-            "video_summary": "The video titled 'Teal'c coffee first time' shows a scene where several individuals are reacting to a beverage, presumably coffee, that Teal'c is trying for the first time. Key moments include: A person off-screen remarking, 'Wow this coffee's great'; another asking if it's 'cinnamon chicory tea oak'; and Teal'c reacting strongly to the taste or temperature, stating 'isn't that hot' indicating he finds it very warm.",
-            "details": {
-                "00:00:00": "Someone remarks, 'Wow this coffee's great I was just thinking that yeah is that cinnamon chicory tea oak'",
-                "00:00:11": "Teal'c takes a large gulp from a black mug",
-                "00:00:24": "Teal'c reacts strongly, someone asks 'isn't that hot'",
-                "00:00:26": "Someone agrees, 'extremely'"
-            }
-        }
-    else:
-        return {"error": "Invalid or unrecognized YouTube URL.", "url": url}
-
-# --- END YOUTUBE TOOL ---
-
 API_KEY = os.getenv("GEMINI_API_KEY")
-HF_API_TOKEN = os.getenv("HF_SPACE_TOKEN") # Kept for potential future HF uses, but not for describe_image
+HF_API_TOKEN = os.getenv("HF_SPACE_TOKEN")  # Kept for potential future HF uses, but not for describe_image
 GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
 
-# Update the tools list (removed describe_image and arvix_search)
+# Updated tools list without Youtube
 tools = [
     multiply, add, subtract, divide, modulus,
     wiki_search,
     google_web_search,
     read_file_content,
     python_interpreter,
-    Youtube,
 ]
 
 with open("prompt.txt", "r", encoding="utf-8") as f:
@@ -206,52 +171,8 @@ def build_graph(provider: str = "gemini"):
 
     def assistant(state: MessagesState):
         messages_to_send = [sys_msg] + state["messages"]
-        
-        # --- IMPORTANT NOTE ON HANDLING BINARY BLOB DATA FOR MULTIMODAL LLMs ---
-        # When read_file_content returns a file_type of "image" or "audio",
-        # the agent should be able to send the actual binary data of that file
-        # as part of the message to the LLM. LangChain's ChatGoogleGenerativeAI
-        # supports this via content parts in HumanMessage.
-        #
-        # For this setup, we're assuming the framework (LangGraph/LangChain)
-        # will correctly handle passing the actual file content when read_file_content
-        # is called and its output indicates a media type.
-        #
-        # A more explicit implementation in the assistant node might look like this
-        # for real binary file handling if the framework doesn't do it implicitly:
-        #
-        # new_messages_to_send = []
-        # for msg in state["messages"]:
-        #    if isinstance(msg, HumanMessage) and msg.tool_calls:
-        #      # If a tool call to read_file_content happened in the previous turn
-        #      # and it returned a media type, we might need to get the file data
-        #      # and append it to the message parts. This logic is complex and
-        #      # depends heavily on how tool outputs are structured and passed.
-        #      # For simplicity in this template, we assume direct handling by the LLM
-        #      # if the tool output indicates media, and the file itself is accessible
-        #      # via the environment.
-        #      pass # Keep original message, tool output will follow
-        #    elif isinstance(msg, HumanMessage) and any(part.get("file_type") in ["image", "audio"] for part in msg.content if isinstance(part, dict)):
-        #      # This is a conceptual example for if the HumanMessage itself contains file data
-        #      # or a reference that needs to be resolved into data.
-        #      # You'd need to load the actual file bytes here.
-        #      # e.g., if msg.content was like: [{"type": "file_reference", "file_path": "image.png"}]
-        #      # with open(msg.content[0]["file_path"], "rb") as f:
-        #      #   file_bytes = f.read()
-        #      # new_messages_to_send.append(
-        #      #     HumanMessage(
-        #      #         content=[
-        #      #             {"type": "text", "text": "Here is the media content:"},
-        #      #             {"type": "image_data" if "image" in msg.content[0]["file_type"] else "audio_data", "data": base64.b64encode(file_bytes).decode('utf-8'), "media_type": "image/png" if "image" in msg.content[0]["file_type"] else "audio/mp3"}
-        #      #         ]
-        #      #     )
-        #      # )
-        #    else:
-        #      new_messages_to_send.append(msg)
-        # llm_response = llm_with_tools.invoke([sys_msg] + new_messages_to_send)
-        # --- END IMPORTANT NOTE ---
 
-        llm_response = llm_with_tools.invoke(messages_to_send,{"recursion_limit": 25}) # For now, keep as is, rely on framework
+        llm_response = llm_with_tools.invoke(messages_to_send, {"recursion_limit": 25})
         print(f"LLM Raw Response: {llm_response}")
         return {"messages": [llm_response]}
 
@@ -265,4 +186,4 @@ def build_graph(provider: str = "gemini"):
     return builder.compile()
 
 if __name__ == "__main__":
-    pass
+    pass