AI-Powered Review Assistant (Agent Pipeline Edition)

An advanced, command-line tool that leverages a dual-agent AI pipeline to generate high-quality, context-aware product reviews. This assistant allows users to connect to multiple Large Language Model (LLM) providers, including Google's Gemini, OpenRouter, Ollama, and LM Studio, to create nuanced and human-like reviews.

---

Project Overview

This repository contains the full source code and documentation for the AI Review Assistant. For a detailed breakdown of the product vision, technical architecture, and future roadmap, please see the official project documents:

• Product Requirements Document (PRD)
• System Flowchart & Legend

---

Key Features

• Dual-Agent Pipeline: Utilizes a "Writer" agent to generate an initial draft and a separate "Editor" agent to refine, humanize, and improve the final output.
• Multi-Provider Support: Easily configure the application to use different LLMs for the writer and editor roles. Supports cloud-based models (Gemini, OpenRouter) and locally-hosted models (Ollama, LM Studio).
• Context-Aware Generation: Can scrape a product's URL to gather real-time product information and analyze past reviews to generate new reviews that are consistent in tone and style.
• Simple CLI Interface: A straightforward menu-driven interface for writing reviews, listing past reviews, and managing settings.
• Configurable & Extensible: All settings, including API keys and model endpoints, are managed in a simple config.json file.

---

How It Works

The application operates on a sophisticated, agent-based pipeline architecture, from user input to final review generation. The core of the system is the AI pipeline, which processes user notes and external data through a sequence of specialized agents. For a detailed explanation, please see the PRD.

Application Flowchart

The following diagram illustrates the complete workflow of the application.

---

Setup and Installation

1. Clone the Repository:

   git clone [https://github.com/your-username/ai-review-assistant.git](https://github.com/your-username/ai-review-assistant.git)
   cd ai-review-assistant

2. Create a Python Environment: It's recommended to use a virtual environment.

   python -m venv venv
   source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

3. Install Dependencies: You will need to install the required Python libraries. Create a requirements.txt file with the following content:

   google-generativeai
   

   Then, install the dependencies:

   pip install -r requirements.txt

   (Note: This project assumes you have a separate fetcher.py script for URL scraping, which may have its own dependencies like BeautifulSoup or requests.)

4. Configure the Application: When you first run the application, a config.json file will be created automatically. Open this file and edit it to select your desired LLM providers and add your API keys.

---

Usage

1. Run the Application: From the root directory, run the main script:

   python main.py

2. Use the Menu:
   • 1. Write a new review: Provide a product name, rating, and notes to generate a review using the standard pipeline.
   • 2. Write a new review with product context: Provide a product URL in addition to your notes to generate a more detailed and context-aware review.
   • 3. List past reviews: View a list of all reviews you have saved in the reviews/ directory.
   • 4. Settings: Configure which LLM provider to use for the "Writer" and "Editor" agents and enter any required API keys.
   • 5. Exit: Close the application.

---

Configuration

The config.json file is the control center for the AI pipeline.

{
  "pipeline": {
    "writer_provider": "gemini",
    "editor_provider": "lmstudio"
  },
  "providers": {
    "gemini": {
      "api_key": "YOUR_API_KEY_HERE"
    },
    "ollama": {
      "endpoint": "http://localhost:11434/api/generate"
    },
    "lmstudio": {
      "endpoint": "http://localhost:1234/v1/chat/completions"
    },
    "openrouter": {
      "api_key": "YOUR_API_KEY_HERE",
      "model": "openrouter/auto"
    }
  }
}

• writer_provider / editor_provider: Can be set to "gemini", "openrouter", "ollama", or "lmstudio".
• api_key: Replace with your actual API key for Gemini or OpenRouter when those providers are used.
• model: Specify the desired OpenRouter model (default openrouter/auto).
• endpoint: Ensure the endpoints for Ollama or LM Studio match your local server configuration.

---

Future Roadmap

Based on the Product Requirements Document, planned future enhancements include:

• Internalize Web Scraping: Integrate the fetcher.py logic directly into the application.
• "Fact-Checker" Agent: Add a third agent to the pipeline to verify the generated review against the scraped product data.
• Tone & Sentiment Control: Allow users to specify the desired tone (e.g., enthusiastic, critical) of the review.
• Batch Generation: Enable review generation for a list of products from a CSV file.
• GUI/Web Interface: Develop a graphical or web-based user interface for broader accessibility.

---

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change. Please make sure to update tests as appropriate.