Plugins Reference

Jac extends its capabilities through a plugin ecosystem. These plugins provide AI integration, full-stack web development, and cloud-native deployment features.

Core Plugins

Plugin	Description	Install
byLLM	AI/LLM integration with Meaning-Typed Programming	`pip install byllm`
jac-client	Full-stack web development with React-style components	`pip install jac-client`
jac-scale	Cloud-native deployment and scaling	`pip install jac-scale`

Quick Installation

Install all plugins via the jaseci meta-package:

pip install jaseci

This bundles jaclang with all official plugins. Or install individually:

pip install byllm        # AI integration
pip install jac-client   # Frontend development
pip install jac-scale    # Production deployment

Plugin Overview

byLLM

Implements Meaning-Typed Programming (MTP) for seamless AI integration:

import from byllm.lib { Model }

glob llm = Model(model_name="gpt-4o");

"""Translate text to the target language."""
def translate(text: str, language: str) -> str by llm();

Full byLLM Reference →

jac-client

Build React-style web applications in Jac:

cl {
    def:pub Counter() -> JsxElement {
        has count: int = 0;

        return <div>
            <p>Count: {count}</p>
            <button onClick={lambda -> None { count = count + 1; }}>
                Increment
            </button>
        </div>;
    }
}

Full jac-client Reference →

jac-scale

Deploy and scale Jac applications:

# Start API server
jac start app.jac

# Deploy to Kubernetes
jac scale deploy --replicas 3

Full jac-scale Reference →

Plugin Architecture

Jac uses a pluggy-based plugin system that allows extending the compiler and runtime:

Compiler plugins - Transform Jac code during compilation
Runtime plugins - Add runtime capabilities (APIs, storage, etc.)
Bundler plugins - Customize build and bundling behavior

Creating Plugins

See the Contributing Guide for information on creating custom plugins.