updated Readme

README.md

@@ -10,4 +10,84 @@ pinned: false
 short_description: Gives Carbon Footprint Score for products description or url
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# Enviromental Impact Analyzer
+This project, "Environmental Impact Analyzer," is a Gradio-based application designed to assess the environmental footprint of products. Users can input either a product description or a URL to a product page, and the application will extract relevant information, calculate its carbon footprint, generate an environmental sustainability score, and provide actionable recommendations for improvement.
+
+--- 
+
+## Features:
+- **Product Information Extraction:** Automatically extracts key environmental factors from text descriptions or web pages (via URL scraping).
+- **Carbon Footprint Calculation:** Integrates with the Climatiq API to estimate the carbon emissions associated with the product's manufacturing and transportation.
+- **Environmental Scoring:** Leverages advanced Large Language Models (LLMs) to generate a comprehensive sustainability score (0-100) based on multiple environmental criteria.
+- **Improvement Recommendations:** Provides specific and actionable suggestions to reduce the product's environmental impact.
+**Intuitive Gradio Interface:** Offers an easy-to-use web interface for seamless interaction.
+
+---
+
+## How It Works
+The core of the Environmental Impact Analyzer is built using **LangGraph**, enabling a stateful, multi-step agentic workflow:
+
+1.  **Input Classification:** Determines if the user input is a direct product description or a URL.
+2.  **Information Extraction:**
+    * If a URL, it **scrapes** the product information from the web page.
+    * If a description, it directly processes the provided text.
+    * An LLM (`deepseek-ai/DeepSeek-R1`) is then used to **extract structured environmental data** (e.g., material composition, manufacturing location, weight, transport distance, recyclability).
+3.  **Carbon Footprint Calculation:** The extracted data is sent to the **Climatiq API** to calculate the estimated carbon footprint (in kg CO2e) based on factors like materials and transportation.
+4.  **Environmental Scoring & Recommendations:**
+    * An LLM (`meta-llama/Meta-Llama-3.1-70B-Instruct`) evaluates the extracted data and carbon footprint.
+    * It generates an **overall environmental score** (0-100), considering factors like carbon emissions, material sustainability, manufacturing practices, transport, longevity, and end-of-life. The scoring is calibrated to recognize positive environmental efforts.
+    * It also provides a **category breakdown** of the score and **actionable recommendations** for improving the product's environmental impact.
+
+**Note:** You can also change the Models for your liking to increse Speed or more accuracy.but make sure hyperbolic supports it.
+
+--- 
+
+## Tech Stack
+
+| Category            | Technology                                 | Purpose                                                              |
+| :------------------ | :----------------------------------------- | :------------------------------------------------------------------- |
+| **Framework** | Gradio                                     | Building the interactive web user interface.                         |
+| **Agentic Workflow**| LangGraph                                  | Orchestrating multi-step agent logic and state management.           |
+| **LLM Integration** | Hugging Face Inference Client              | Interacting with Large Language Models hosted on Hugging Face Hub.   |
+| **Core LLM** | `deepseek-ai/DeepSeek-R1`                  | Generates environmental scores, category breakdowns, and recommendations. |
+| **External APIs** | Climatiq API                               | Calculates carbon footprint based on product data.                   |
+| **Web Scraping** | `requests`, `BeautifulSoup` (via `scrapper.py`) | Fetching and parsing HTML content from product URLs.                 |
+| **Prompting** | `langchain.prompts.ChatPromptTemplate`     | Structuring and templating LLM prompts for consistent input.         |
+| **Environment Mgt.**| `python-dotenv`                            | Loading environment variables from `.env` files for secure key management. |
+| **Data Typing** | `typing`, `typing_extensions` (`TypedDict`) | Enhancing code readability and maintainability with type hints.      |
+
+---
+
+## Setup and Installation
+
+To run this project locally, follow these steps:
+
+### 1. Clone the Repository
+
+```bash
+git clone [https://huggingface.co/spaces/Agents-MCP-Hackathon/Environmental-Impact-Analyzer](https://huggingface.co/spaces/Agents-MCP-Hackathon/Environmental-Impact-Analyzer)
+cd Environmental-Impact-Analyzer
+```
+
+### 2. Create a Virtual Environment (Recommended)
+```bash
+python -m venv .venv
+# On Windows
+.venv\Scripts\activate
+# On macOS/Linux
+source .venv/bin/activate
+```
+### 3. Install Dependencies
+```bash
+pip install -r requirements.txt
+```
+### 4. 4. Configure Environment Variables
+Create a .env file in the root directory of your project and add the following:
+```
+HF_API_KEY="hf_YOUR_HUGGING_FACE_ACCESS_TOKEN"
+CLIMATIQ_API_KEY="YOUR_CLIMATIQ_API_KEY"
+```
+### 5. Run the Application
+```bash
+python app.py
+```

app.py

@@ -84,7 +84,7 @@ def create_gradio_interface():
             with gr.Column(scale=2):
                 # Input section
                 product_input = gr.Textbox(
-                    label="Product Description",
+                    label="Product Description/URL",
                     placeholder="e.g., Organic cotton t-shirt manufactured in India, packaged in recyclable materials...",
                     lines=4,
                     max_lines=8
@@ -101,7 +101,7 @@ def create_gradio_interface():
                 with gr.Group():
                     score_output = gr.Markdown(
                         label="Environmental Score",
-                        value="Enter a product description and click analyze to see results."
+                        value="Enter a product description/URL and click analyze to see results."
                     )
                     
                     status_output = gr.Markdown(
@@ -141,6 +141,11 @@ def create_gradio_interface():
             inputs=[product_input],
             label="Example Products to Analyze"
         )
+        gr.Markdown("""
+        **Note**: add the Following Secret Key to your environment variables:
+                    - `HF_API_KEY` (Hugging Face API token) to access private models or higher rate limits.
+                    - `CLIMATIQ_API_KEY` (Climatiq API token) for carbon footprint calculations.
+                    """)
     
     return interface