PromptPilot 🚀

A Django-based AI prompt management and chat platform, blending modern APIs, classic templates, and machine learning moderation — built over nearly two months of daily development.

---

Overview

PromptPilot is a full-stack Django project integrating Django REST Framework for backend APIs with traditional Django templates for essential user flows.
It connects to Ollama’s Mistral model for AI-powered chat, supports both browser and terminal interaction, and includes an scikit-learn-powered comment moderation system.

---

Features

• 5 DRF API endpoints for prompt management, message storage, and more.
• Traditional Django templates for:
  • Login / Logout
  • Signup
  • Home
  • Password Reset
• JWT Authentication via django-simple-jwt for secure token-based API access.
• Machine Learning Auto-Moderation with scikit-learn to flag inappropriate comments.
• Secure Secrets Management with python-dotenv.
• Integration with Ollama Mistral API:
  • Communicate with Mistral directly from the Bash terminal in VS Code.
  • Store AI responses in a dedicated API endpoint.
• Token Auto-Refresh Script to keep .env credentials up-to-date.
• Code Style Enforcement with black formatter and crispy forms.
• 14 Automated Tests for reliability and regression prevention.

---

Screenshots

🖥️ Home Page

💬 Terminal Chat with Mistral

🔌 API Playground

🔑 Password Reset Flow

📜 Chat Message History

---

Project Structure

promptpilot/
├── prompts/          # Core prompt models, serializers, and views
├── users/            # User authentication & profile management
├── playgrounds/      # AI prompt execution and logging
├── chatmessages/     # Stores AI conversation history
├── scripts/          # Utility scripts (e.g., token refresh)
├── static/           # Static files
├── templates/        # HTML templates
├── utils/            # Helper functions/utilities
├── .dockerignore
├── .gitignore
├── .env.example
├── README.md
├── requirements.txt
├── manage.py
└── ...

---

Setup & Installation

1. Clone the repository

git clone https://github.com/pablo727/promptpilot
cd promptpilot

2. Create & activate virtual environment

python -m venv .venv
source .venv/bin/activate  # Linux / macOS
.venv\Scripts\activate     # Windows

3. Install dependencies

pip install -r requirements.txt

4. Set up environment variables

cp .env.example .env
# Fill in your secrets and API keys

5. Run migrations

python manage.py migrate

6. Run the server

python manage.py runserver

Ollama Terminal Chat

You can start a real-time chat with the Mistral model directly from your terminal in VS Code. Responses are stored automatically via the DRF API for future retrieval.

Testing

Run all tests:

python manage.py test

There are 14 automated tests covering models, views, serializers, and integration points.

Machine Learning Auto-Moderation

The project includes a flag_classifier.py built with scikit-learn to detect and flag potentially inappropriate comments. You can train the model via:

python manage.py train_flag_model

License

This project is licensed under the MIT License. Collaborations are welcome!