@shoprag/genius

The Genius System is a powerful, flexible AI-driven conversational system designed to facilitate the creation and deployment of specialized AI assistants, or "Geniuses." While it comes with two default Geniuses—bitgenius and poetic-bitgenius—the true strength of this system lies in its ability to support any type of Genius you can imagine. Whether you're building an assistant for blockchain, poetry, or any other domain, @shoprag/genius provides the infrastructure to bring your vision to life.

This system is built with developers in mind, offering a robust set of features, including support for multiple AI providers, customizable prompt data, and a fully configurable environment. We welcome contributions to expand the ecosystem with new Geniuses and enhancements—PRs are always welcome!

Features

• Multiple AI Providers: Seamlessly switch between AI providers like OpenAI and Ollama, with support for more via custom implementations.
• Customizable Geniuses: Define your own Genius with custom system prompts and context prefixes using JSON configuration files.
• User Authentication: Secure user registration and login with email verification, powered by Sendgrid.
• Conversational Management: Create, manage, and delete conversations with ease.
• Streaming Support: Real-time streaming of AI responses for a dynamic user experience.
• Configurable Logging: Tailor logging levels to your needs, from debug to critical.
• Database Flexibility: Use SQLite by default or configure for MySQL or other databases via Knex.
• Rate Limiting: Protect your server with built-in rate limiting.
• Input Sanitization: Automatic sanitization of inputs to prevent XSS attacks.

Installation

To get started, install @shoprag/genius globally using npm or yarn:

npm install -g @shoprag/genius

or

yarn global add @shoprag/genius

Configuration

@shoprag/genius can be configured using command-line flags, environment variables, or a combination of both. Command-line flags take precedence over environment variables.

Essential Configuration Options

• --port <port>: The port on which the server will listen (default: 3000).
• --ai-provider <provider>: The AI provider to use (openai or ollama).
• --openai-api-key <key>: API key for OpenAI (required if using OpenAI).
• --openai-base-url <url>: Base URL for OpenAI API (default: https://api.openai.com/v1).
• --openai-model <model>: OpenAI model to use (default: gpt-4o).
• --ollama-base-url <url>: Base URL for Ollama API (default: http://localhost:11434).
• --ollama-model <model>: Ollama model to use (default: llama2).
• --universe-url <url>: URL of the Universe API.
• --universe-bearer <token>: Bearer token for Universe API authentication.
• --jwt-secret <secret>: Secret key for JWT authentication.
• --sendgrid-api-key <key>: API key for Sendgrid email services.
• --sendgrid-from-email <email>: Verified sender email for Sendgrid (e.g., [email protected]).
• --genius <genius>: The Genius preset to use (bitgenius, poetic-bitgenius, or custom).
• --custom-genius-json <path>: Path to a custom Genius JSON file (required if --genius=custom).

Database Configuration

• --db-type <type>: Database type (sqlite or mysql, default: sqlite).
• --db-file <file>: Path to SQLite database file (default: ./db.sqlite).
• --db-host <host>: Database host (for MySQL).
• --db-user <user>: Database user (for MySQL).
• --db-password <password>: Database password (for MySQL).
• --db-name <name>: Database name (for MySQL).

Logging and Miscellaneous

• --log-level <level>: Logging level (debug, info, critical, default: info).

Setup Wizard

If essential configurations (like API keys or the JWT secret) are missing, a setup wizard will guide you through providing the necessary details. The wizard will create a .env file in the current working directory with your inputs.

Example .env File

PORT=3000
AI_PROVIDER=openai
OPENAI_API_KEY=your-openai-key
SENDGRID_API_KEY=your-sendgrid-key
[email protected]
JWT_SECRET=your-secret-key

Usage

Starting the Server

To start the Genius server, run:

genius

If you haven’t configured the system yet, the setup wizard will prompt you for the required information.

Registering a User

To register a new user, send a POST request to /register with the user’s email and password:

curl -X POST http://localhost:3000/register -H "Content-Type: application/json" -d '{"email": "[email protected]", "password": "strongpassword"}'

The user will receive a verification email with a code that must be used to verify their email.

Verifying Email

To verify the email, send a POST request to /verify with the email and verification code:

curl -X POST http://localhost:3000/verify -H "Content-Type: application/json" -d '{"email": "[email protected]", "code": "ABC123"}'

Logging In

Once verified, log in by sending a POST request to /login:

curl -X POST http://localhost:3000/login -H "Content-Type: application/json" -d '{"email": "[email protected]", "password": "strongpassword"}'

You will receive a JWT token in the response, which must be included in the Authorization header for authenticated requests (e.g., Authorization: Bearer <token>).

Creating a Conversation

To create a new conversation, send a POST request to /convo with the conversation title:

curl -X POST http://localhost:3000/convo -H "Authorization: Bearer <token>" -H "Content-Type: application/json" -d '{"title": "My First Convo"}'

Sending a Message

To send a message in a conversation, send a POST request to /send/<convo_id>:

curl -X POST http://localhost:3000/send/1 -H "Authorization: Bearer <token>" -H "Content-Type: application/json" -d '{"message": "Hello, Genius!", "stream": true}'

Set "stream": true to receive real-time streaming responses via Server-Sent Events (SSE).

Deleting a Conversation

To delete a conversation, send a DELETE request to /convo/<convo_id>:

curl -X DELETE http://localhost:3000/convo/1 -H "Authorization: Bearer <token>"

To delete all conversations, use /convo/all:

curl -X DELETE http://localhost:3000/convo/all -H "Authorization: Bearer <token>"

API Documentation

Below is a comprehensive list of all available API endpoints, including their parameters and expected responses.

POST /register

• Description: Registers a new user.
• Parameters:
  • email (string, required): User’s email.
  • password (string, required): User’s password.
• Response:
  • 200: { "message": "User registered, please check your email for verification code" }
  • 400: { "message": "Error message" }

POST /verify

• Description: Verifies a user’s email.
• Parameters:
  • email (string, required): User’s email.
  • code (string, required): Verification code from email.
• Response:
  • 200: { "message": "Email verified successfully" }
  • 400: { "message": "Invalid or expired code" }

POST /resend-verification

• Description: Resends the verification code to the user’s email.
• Parameters:
  • email (string, required): User’s email.
• Response:
  • 200: { "message": "Verification code resent" }
  • 400: { "message": "Error message" }

POST /login

• Description: Logs in a user and returns a JWT token.
• Parameters:
  • email (string, required): User’s email.
  • password (string, required): User’s password.
• Response:
  • 200: { "message": "Logged in successfully", "token": "jwt_token" }
  • 400: { "message": "Invalid credentials" }

POST /convo

• Description: Creates a new conversation.
• Parameters:
  • title (string, required): Conversation title.
• Headers:
  • Authorization: Bearer <token>
• Response:
  • 200: { "message": "Conversation created successfully", "id": "convo_id" }
  • 401: { "message": "Token required" }

GET /convos

• Description: Retrieves all conversations for the authenticated user.
• Headers:
  • Authorization: Bearer <token>
• Response:
  • 200: Array of conversation objects (e.g., [{ "id": "1", "title": "My First Convo" }])
  • 401: { "message": "Token required" }

GET /convo/:id

• Description: Retrieves messages for a specific conversation.
• Parameters:
  • id (string, required): Conversation ID.
• Headers:
  • Authorization: Bearer <token>
• Response:
  • 200: Array of message objects (e.g., [{ "sender": "user", "content": "Hello" }, { "sender": "genius", "content": "Hi there!" }])
  • 404: { "message": "Conversation not found" }

POST /send/:id

• Description: Sends a message in a conversation and receives a response from the Genius.
• Parameters:
  • message (string, required): User’s message.
  • subject (string, optional): Subject for context.
  • exclusive (boolean, optional): Use exclusive context (default: false).
  • fast (boolean, optional): Use a faster but less accurate model (default: false).
  • responseLength (number, optional): Desired response length in characters.
  • stream (boolean, optional): Enable streaming response (default: false).
• Headers:
  • Authorization: Bearer <token>
• Response:
  • 200: { "response": "Genius response" } (non-streaming) or SSE events (streaming).
  • 404: { "message": "Conversation not found" }

DELETE /convo/:id

• Description: Deletes a conversation or all conversations if id is "all".
• Parameters:
  • id (string, required): Conversation ID or "all".
• Headers:
  • Authorization: Bearer <token>
• Response:
  • 200: { "message": "Conversation deleted successfully" }
  • 404: { "message": "Conversation not found" }

Contributing

We welcome contributions to @shoprag/genius! Whether you’re adding a new Genius, improving existing features, or fixing bugs, your input is valuable.

How to Contribute

1. Fork the Repository: Start by forking the project on GitHub.
2. Create a Branch: Create a new branch for your feature or bugfix (e.g., feature/new-genius).
3. Make Changes: Implement your changes, ensuring they align with the project’s coding standards.
4. Test: Thoroughly test your changes to ensure they work as expected.
5. Submit a Pull Request: Open a PR with a clear description of your changes and why they’re needed.

Adding a New Genius

To add a new Genius:

• Create a JSON file with the following structure:

  {
    "systemPrompt": "You are a helpful assistant specialized in [domain].",
    "contextPrefix": "Providing context for [domain]: "
  }

• Specify the path to this file using --custom-genius-json when starting the server.
• Submit a PR with your new Genius file and any related code changes.

We can’t wait to see what amazing Geniuses you’ll bring to the world!

License

This project is licensed under the MIT License.