Comment Review AI

An agentic application that analyzes user comments for profanity, sentiment, and relevance to movies or books. It fetches additional information from OMDb and provides a simple API and UI to interact with the system.

Project Structure

• src/: Contains the core application logic.
  • main.py: The FastAPI application.
  • agent.py: The Pydantic AI agent for comment analysis.
  • database.py: A simple in-memory database to store comment analysis results.
• gradio_app.py: The Gradio web interface.
• mcp_client.py: The Master Control Program client to test the API.
• requirements.txt: Python dependencies.
• .env.example: Example environment file. Copy to .env and fill in your details.
• start.sh: Script to start the FastAPI server and Gradio app.
• stop.sh: Script to stop the running processes.

Setup

1. Install dependencies:

   Option A: Using uv (Recommended - Faster)

   uv is an extremely fast Python package installer and resolver.

   • Install uv:

     # macOS/Linux
     curl -LsSf https://astral.sh/uv/install.sh | sh
     
     # Windows
     powershell -c "irm https://astral.sh/uv/install.ps1 | iex"
     
     # Or with pip
     pip install uv

   • Install dependencies with uv:

     uv pip install -r requirements.txt

   • Create and activate virtual environment with uv:

     uv venv
     source .venv/bin/activate  # On macOS/Linux
     # or
     .venv\Scripts\activate  # On Windows

   Option B: Using pip (Traditional)

   pip install -r requirements.txt

2. Set up Ollama (Local LLM):

   This project uses Ollama to run a local LLM (llama3.1) for comment analysis.

   • Install Ollama:

     • macOS/Linux:

       curl -fsSL https://ollama.ai/install.sh | sh

     • Windows: Download from https://ollama.ai/download
     • Or visit: https://ollama.ai for other installation methods

   • Pull the llama3.1 model:

     ollama pull llama3.1:latest

   • Start Ollama (if not already running):

     ollama serve

     Ollama will run on http://localhost:11434 by default.

   • Verify Ollama is running:

     curl http://localhost:11434/api/tags

     You should see a list of available models including llama3.1:latest.

   Alternative: Using OpenAI instead of Ollama

   If you prefer to use OpenAI's API instead of a local model, modify src/agent.py:

   # Replace the ollama_model configuration with:
   ai_analyzer = Agent(
       "openai:gpt-4-turbo",  # or "openai:gpt-3.5-turbo"
       output_type=CommentAnalysis,
   )

   And set your OpenAI API key in .env:

   OPENAI_API_KEY=your_openai_api_key_here
   

3. Set up environment variables:

   • Copy .env.example to a new file named .env.
   • Get a free API key from OMDb API.
   • Add your API key to the .env file:

     OMDB_API_KEY=YOUR_API_KEY_HERE
     

Running the Application

• To start everything:

  ./start.sh

  • FastAPI will be available at http://127.0.0.1:8000.
  • Gradio will be available at http://127.0.0.1:7860.

• To stop everything:

  ./stop.sh