New image tool added.

.gitignore

@@ -0,0 +1 @@
+.env

app.py

@@ -4,15 +4,41 @@ import requests
 import inspect
 import pandas as pd
 
+import requests
+from smolagents import OpenAIServerModel
+from smolagents import CodeAgent, DuckDuckGoSearchTool, VisitWebpageTool, FinalAnswerTool, ToolCallingAgent
+
+from modules.tools import image_description
+
 # (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 ------
+
 class BasicAgent:
     def __init__(self):
         print("BasicAgent initialized.")
+        
+        self.model = OpenAIServerModel(
+            model_id="gemini-2.5-flash-preview-05-20",
+            # Google Gemini OpenAI-compatible API base URL
+            api_base="https://generativelanguage.googleapis.com/v1beta/openai/",
+            api_key=GEMINI_API_KEY,
+            )
+        
+        self.agent = CodeAgent (
+            tools = [DuckDuckGoSearchTool(), VisitWebpageTool(),
+                     image_description,
+                    #  my_cumtom_tool_class()
+                     ],
+                     model = self.model,
+                     max_steps = 10,
+                     verbosity_level = 2, 
+                     additional_authorized_imports = ["datetime"]
+                     )
+
     def __call__(self, question: str) -> str:
         print(f"Agent received question (first 50 chars): {question[:50]}...")
         fixed_answer = "This is a default answer."

modules/prompts.py

@@ -0,0 +1,14 @@
+image_description_prompt = \
+"""
+Analyze this image in detail to help answer the following question: '{question}'
+
+Provide a structured description with the following sections:
+
+1.  **Key Objects & Entities:** List all primary objects, people, text, symbols, and entities. Be specific about quantities, colors, and spatial relationships.
+2.  **Visual Context & Setting:** Describe the environment, background, layout, and any relevant aesthetic or stylistic elements (e.g., "a vintage infographic," "a modern UI screenshot").
+3.  **Data Presentation (if applicable):** If the image contains data, describe the chart type, axes labels, units, data trends, and approximate values. Ignore this section if not applicable.
+4.  **Textual Content:** Transcribe all visible text exactly as it appears, including labels, headings, and fine print.
+5.  **Relevance to Question:** Based on the above, briefly state which visual elements are most directly relevant to answering the user's question.
+
+Your description must be factual, concise, and objective. Avoid speculation and interpretation not supported by visual evidence.
+"""

modules/tools.py

@@ -0,0 +1,26 @@
+import requests
+from google import genai
+from google.genai import types
+from smolagents import tool
+from modules.prompts import image_description_prompt
+
+@tool
+def image_description(question: str, image_url: str) -> str:
+    """
+    Recieves an image url and analyze it first, then return a what is needed based on prompt.
+    Args:
+        question: a question about given image
+        image_url: the url of image which should be considered as a reference for answering the question.
+    """
+    image_bytes = requests.get(image_url).content
+    image = types.Part.from_bytes(
+        data=image_bytes, mime_type="image/jpeg"
+        )
+    
+    client = genai.Client()
+    response = client.models.generate_content(
+        model="gemini-2.5-flash",
+        contents=[image_description_prompt.format(question=question), image],
+        )
+
+    return response.text

requirements.txt

@@ -1,2 +1,6 @@
 gradio
-requests
+gradio[oauth]
+requests
+smolagents
+google-api-python-client
+google-genai

tests/genai.ipynb

@@ -0,0 +1,104 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "ed655f3a",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from google import genai\n",
+    "\n",
+    "# The client gets the API key from the environment variable `GEMINI_API_KEY`.\n",
+    "client = genai.Client()\n",
+    "\n",
+    "response = client.models.generate_content(\n",
+    "    model=\"gemini-2.5-flash\", contents=\"write a poem.\"\n",
+    ")\n",
+    "print(response.text)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "133682b5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from google import genai\n",
+    "from google.genai import types\n",
+    "import requests\n",
+    "\n",
+    "@tool\n",
+    "def image_description(image_url: str, image_description_prompt: str) -> str:\n",
+    "    \"\"\"\n",
+    "    Recieves an image url and analyze it first, then return a what is needed based on prompt.\n",
+    "    Args:\n",
+    "        image_url: the url of image which should be described.\n",
+    "        image_description_prompt: a reference prompt which control model output.\n",
+    "    \"\"\"\n",
+    "    image_bytes = requests.get(image_url).content\n",
+    "    image = types.Part.from_bytes(\n",
+    "        data=image_bytes, mime_type=\"image/jpeg\"\n",
+    "        )\n",
+    "    \n",
+    "    client = genai.Client()\n",
+    "    response = client.models.generate_content(\n",
+    "        model=\"gemini-2.5-flash\",\n",
+    "        contents=[image_description_prompt, image],\n",
+    "        )\n",
+    "\n",
+    "    return response.text"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "bc350e9e",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from google import genai\n",
+    "from google.genai import types\n",
+    "\n",
+    "import requests\n",
+    "\n",
+    "image_path = \"https://goo.gle/instrument-img\"\n",
+    "image_bytes = requests.get(image_path).content\n",
+    "image = types.Part.from_bytes(\n",
+    "  data=image_bytes, mime_type=\"image/jpeg\"\n",
+    ")\n",
+    "\n",
+    "client = genai.Client()\n",
+    "\n",
+    "response = client.models.generate_content(\n",
+    "    model=\"gemini-2.5-flash\",\n",
+    "    contents=[\"What is this image?\", image],\n",
+    ")\n",
+    "\n",
+    "print(response.text)"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": ".venv",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.12.3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}