KnowPilot is an agent system designed for private domains, combining human-centered design with knowledge augmentation. It features a modern frontend, robust backend, and seamless AI integration for document management, knowledge retrieval, and collaborative writing.

• Install Docker
• Clone this repository:

git clone https://github.com/XeeKee/KnowPilot.git

1. Prepare Environment Variables

   cp docker/env.example docker/.env
   # Edit docker/.env and fill in the following:
   # - SECRET_KEY (application secret, use a long random string)
   # - POSTGRES_PASSWORD (PostgreSQL database password)
   # - MODEL_API_KEY (Volcano Engine API key)
   # - SEARCH_API_KEY (Search API key, optional)

2. Start Services

   ./start.sh
   # Or start manually
   docker compose -f docker/docker-compose.yml up -d --build

3. Access & Verification

   • Frontend: http://localhost
   • Backend health check: http://localhost:5000/api/health
   • View logs (for troubleshooting):

     docker compose -f docker/docker-compose.yml logs -f backend
     docker compose -f docker/docker-compose.yml logs -f db
     docker compose -f docker/docker-compose.yml logs -f frontend

4. Stop Services

   ./stop.sh
   # Or stop manually
   docker compose -f docker/docker-compose.yml down

• docker/docker-compose.yml
  • Uses named volume postgres_data for DB
  • Backend reads config from environment variables (not config files)
• docker/config.docker.yaml
  • Docker-specific config, uses environment variable placeholders
• docker/.env
  • Environment variable config file
• docker/env.example
  • Example environment file, provides configuration guidance
• front_end_re/
  • Frontend static assets, mounted read-only in Nginx container
• src_re/
  • Flask backend and business logic

1. Create a Python virtual environment
2. Install dependencies: pip install -r requirements.txt
3. Set environment variables or edit src_re/config.yaml
4. Start: python src_re/app.py

• Chat Interface (chat.html): Topic input, outline generation, article creation
• Private Library Management (private_library.js): Document upload, vector storage, knowledge retrieval
• Article Generation (generate_article.js): Section generation, editing, export
• Outline Management (send_topic.js): Outline generation, feedback, sorting

• AI Model Integration: Supports Volcano Engine DeepSeek models
• Vector Database: ChromaDB integration for document vector storage
• Search Enhancement: External search engine integration for real-time information
• Session Management: User session state management, multi-user concurrency

• Frontend: Vanilla JavaScript + HTML5 + CSS3
• Backend: Flask + SQLAlchemy + PostgreSQL
• AI/ML: LangChain + Sentence Transformers + ChromaDB
• Deployment: Docker + Docker Compose + Nginx
• Database: PostgreSQL + ChromaDB vector storage

This project is licensed under the MIT License. See the LICENSE file for details.