changes in tools

app.py

@@ -10,16 +10,24 @@ from langgraph.prebuilt import ToolNode, tools_condition
 from langgraph.graph.message import add_messages
 from typing import TypedDict, Annotated
 from tools import (
-    image_recognition_tool,
-    download_file,
-    reverse_text_tool,
     serp_search_tool,
     python_execution_tool,
-    video_analysis_tool,
+    reverse_text_tool,
     audio_processing_tool,
+    video_analysis_tool,
+    image_recognition_tool,
     file_type_detection_tool,
     read_file_tool,
-    wiki_search_tool
+    code_execution_tool,
+    math_calculation_tool,
+    wiki_search_tool,
+    python_repl_tool,
+    save_and_read_file_tool,
+    download_file_from_url_tool,
+    extract_text_from_image_tool,
+    analyze_csv_file_tool,
+    analyze_excel_file_tool,
+    download_file,
 )
 import re
 
@@ -28,18 +36,25 @@ import re
 DEFAULT_API_URL = "https://agents-course-unit4-scoring.hf.space"
 
 # Setting up the llm
-llm = ChatOpenAI(model="gpt-3.5-turbo", temperature=0)
+llm = ChatOpenAI(model="gpt-4o", temperature=0)
 tools = [
     serp_search_tool,
     wiki_search_tool,
-    download_file,
-    image_recognition_tool,
-    reverse_text_tool,
     python_execution_tool,
-    video_analysis_tool,
+    reverse_text_tool,
     audio_processing_tool,
+    video_analysis_tool,
+    image_recognition_tool,
     file_type_detection_tool,
-    read_file_tool
+    read_file_tool,
+    code_execution_tool,
+    math_calculation_tool,
+    python_repl_tool,
+    save_and_read_file_tool,
+    download_file_from_url_tool,
+    extract_text_from_image_tool,
+    analyze_csv_file_tool,
+    analyze_excel_file_tool,
 ]
 chat_with_tools = llm.bind_tools(tools)
 
@@ -150,53 +165,11 @@ class AgentWrapper:
 def assistant(state: MyAgent):
     # Add system message to instruct the agent to use the tool
     system_message = SystemMessage(content="""
-You are a general AI assistant.
-
-**Only output the answer. Do not include any explanation, context, or restate the question.**
-If the answer is a number, output only the number.
-If the answer is a string, output only the string.
-If the answer is a list, output only a comma-separated list.
-Do not use any prefixes, suffixes, or extra words.
-
-If you're asked for a number, don't use commas or units like $ or %, unless specified.
-If you're asked for a string, don't use articles or abbreviations (e.g. for cities), and write digits in plain text unless told otherwise.
-
-IMPORTANT: You have a maximum of 3 tool uses per question. After 3 tool uses, you MUST provide your best answer based on available information.
-
-DECISION MAKING:
-1. First, try to answer from your knowledge if it's a general fact.
-2. If you need specific, current, or detailed information, use serp_search_tool ONCE.
-3. If the question looks reversed (starts with a period), use reverse_text_tool ONCE first.
-4. For factual or historical questions, use wiki_search_tool ONCE.
-5. For file-based questions, use the appropriate file tool.
-6. After using a tool, analyze the result and provide your final answer.
-7. Do NOT cycle between tools unnecessarily.
-
-Tool Use Guidelines:
-1. Do **not** use any tools outside of the provided tools list.
-2. Always use **only one tool at a time** in each step of your execution.
-3. You have a MAXIMUM of 3 tool uses per question.
-4. For web searches and current information, use **serp_search_tool** (15s timeout).
-5. For factual or historical questions, use **wiki_search_tool**.
-6. If the question looks reversed (starts with a period or reads backward), first use **reverse_text_tool** to reverse it, then process the question.
-7. For image analysis and description, use **image_recognition_tool** (requires OpenAI API key).
-8. For Python code execution, use **python_execution_tool**.
-9. For video analysis, use **video_analysis_tool**.
-10. For audio processing, use **audio_processing_tool**.
-11. For file type detection, use **file_type_detection_tool**.
-12. For reading file contents, use **read_file_tool**.
-13. File downloading is handled automatically - you don't need to download files manually.
-14. Keep responses concise and efficient.
-15. If you can't find the answer after using 2-3 tools, provide your best estimate based on available information.
-16. NEVER use more than 3 tools for a single question.
-17. After using a tool, provide your final answer immediately.
-
-FILE PROCESSING:
-- Questions may come with attached files (mp3, excel, images, etc.)
-- File processing is automatically handled before the question reaches you
-- If a file is attached, you'll see the file content/analysis in the question
-- Focus on answering based on the enhanced question that includes file information
-- You don't need to process files manually - just use the information provided
+You are a helpful assistant tasked with answering questions using a set of tools. 
+Now, I will ask you a question. Report your thoughts, and finish your answer with the following template: 
+FINAL ANSWER: [YOUR FINAL ANSWER]. 
+YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings. If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise. If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, Apply the rules above for each element (number or string), ensure there is exactly one space after each comma.
+Your answer should only start with "FINAL ANSWER: ", then follows with the answer. 
 """)
 
     # Combine system message with user messages
@@ -410,3 +383,11 @@ if __name__ == "__main__":
 
     print("Launching Gradio Interface for Basic Agent Evaluation...")
     demo.launch(debug=True, share=False)
+
+
+# test
+messages = [HumanMessage(
+    content="Question: How many studio albums were published by Mercedes Sosa between 2000 and 2009 (included)? You can use the latest 2022 version of english wikipedia.")]
+response = my_agent.invoke({"messages": messages})
+print("🎩 Alfred's Response:")
+print(response['messages'][-1].content)

requirements.txt

@@ -17,4 +17,11 @@ langchain-huggingface
 pydub
 google-search-results
 pytesseract
-numpy 
+numpy 
+wikipedia
+langchain_experimental
+uuid
+openpyxl
+xlrd
+ffmpeg-python
+opencv-python

test_agent.py

@@ -0,0 +1,38 @@
+import os
+import requests
+
+
+def fetch_questions():
+    api_url = os.getenv("DEFAULT_API_URL",
+                        "https://agents-course-unit4-scoring.hf.space")
+    questions_url = f"{api_url}/questions"
+    print(f"Fetching questions from: {questions_url}")
+    try:
+        response = requests.get(questions_url, timeout=15)
+        response.raise_for_status()
+        questions_data = response.json()
+        if not questions_data:
+            print("Fetched questions list is empty.")
+            return []
+        print(f"Fetched {len(questions_data)} questions.")
+        return questions_data
+    except Exception as e:
+        print(f"Error fetching questions: {e}")
+        return []
+
+
+def main():
+    questions = fetch_questions()
+    if not questions:
+        print("No questions to display.")
+        return
+    print("\nFirst 20 questions from the API:")
+    for idx, q in enumerate(questions[:20], 1):
+        print(f"\n[{idx}] Task ID: {q.get('task_id')}")
+        print(f"Question: {q.get('question')}")
+        if q.get('file_name'):
+            print(f"Attached file: {q['file_name']}")
+
+
+if __name__ == "__main__":
+    main()

tools.py

@@ -21,7 +21,10 @@ from pydub import AudioSegment
 import pandas as pd
 from PIL import Image
 from langchain_community.document_loaders import WikipediaLoader
-
+from langchain_experimental.tools.python.tool import PythonREPLTool
+import uuid
+import pytesseract
+from urllib.parse import urlparse
 # Load environment variables
 print("Current working directory:", os.getcwd())
 
@@ -546,3 +549,112 @@ wiki_search_tool = Tool(
     func=wiki_search,
     description="Search Wikipedia for a query and return up to 2 results. Use this for factual or historical questions."
 )
+
+python_tool = PythonREPLTool()
+
+python_repl_tool = Tool(
+    name="python_repl_tool",
+    func=python_tool,
+    description="Executes Python code in a REPL environment. Use this for running Python code snippets interactively."
+)
+
+# --- New Tools ---
+
+
+def save_and_read_file(content: str, filename: str = None) -> str:
+    temp_dir = tempfile.gettempdir()
+    if filename is None:
+        temp_file = tempfile.NamedTemporaryFile(delete=False, dir=temp_dir)
+        filepath = temp_file.name
+    else:
+        filepath = os.path.join(temp_dir, filename)
+    with open(filepath, "w") as f:
+        f.write(content)
+    return f"File saved to {filepath}. You can read this file to process its contents."
+
+
+save_and_read_file_tool = Tool(
+    name="save_and_read_file_tool",
+    func=save_and_read_file,
+    description="Save content to a file and return the path. Optionally specify a filename."
+)
+
+
+def download_file_from_url(url: str, filename: str = None) -> str:
+    try:
+        if not filename:
+            path = urlparse(url).path
+            filename = os.path.basename(path)
+            if not filename:
+                filename = f"downloaded_{uuid.uuid4().hex[:8]}"
+        temp_dir = tempfile.gettempdir()
+        filepath = os.path.join(temp_dir, filename)
+        response = requests.get(url, stream=True)
+        response.raise_for_status()
+        with open(filepath, "wb") as f:
+            for chunk in response.iter_content(chunk_size=8192):
+                f.write(chunk)
+        return f"File downloaded to {filepath}. You can read this file to process its contents."
+    except Exception as e:
+        return f"Error downloading file: {str(e)}"
+
+
+download_file_from_url_tool = Tool(
+    name="download_file_from_url_tool",
+    func=download_file_from_url,
+    description="Download a file from a URL and save it to a temporary location. Optionally specify a filename."
+)
+
+
+def extract_text_from_image(image_path: str) -> str:
+    try:
+        image = Image.open(image_path)
+        text = pytesseract.image_to_string(image)
+        return f"Extracted text from image:\n\n{text}"
+    except Exception as e:
+        return f"Error extracting text from image: {str(e)}"
+
+
+extract_text_from_image_tool = Tool(
+    name="extract_text_from_image_tool",
+    func=extract_text_from_image,
+    description="Extract text from an image using OCR (pytesseract)."
+)
+
+
+def analyze_csv_file(file_path: str, query: str) -> str:
+    try:
+        df = pd.read_csv(file_path)
+        result = f"CSV file loaded with {len(df)} rows and {len(df.columns)} columns.\n"
+        result += f"Columns: {', '.join(df.columns)}\n\n"
+        result += "Summary statistics:\n"
+        result += str(df.describe())
+        return result
+    except Exception as e:
+        return f"Error analyzing CSV file: {str(e)}"
+
+
+analyze_csv_file_tool = Tool(
+    name="analyze_csv_file_tool",
+    func=analyze_csv_file,
+    description="Analyze a CSV file using pandas and answer a question about it."
+)
+
+
+def analyze_excel_file(file_path: str, query: str) -> str:
+    try:
+        df = pd.read_excel(file_path)
+        result = f"Excel file loaded with {len(df)} rows and {len(df.columns)} columns.\n"
+        result += f"Columns: {', '.join(df.columns)}\n\n"
+        result += "Summary statistics:\n"
+        result += str(df.describe())
+        return result
+    except Exception as e:
+        return f"Error analyzing Excel file: {str(e)}"
+
+
+analyze_excel_file_tool = Tool(
+    name="analyze_excel_file_tool",
+    func=analyze_excel_file,
+    description="Analyze an Excel file using pandas and answer a question about it."
+)