Getting Started with Jivas

Welcome to the Jivas Agentic Framework! Let's get started with installing the framework and setting up your first project.

Installation

To install Jivas, you'll need to use pip. Run the following commands in your terminal:

pip install jivas
pip install git+https://github.com/TrueSelph/jaseci.git@fast_import_v2#egg=jaclang&subdirectory=jac

Ensure you have Python installed on your system before running these commands.

Once Jivas is installed, you can create a new project using the jvcli command-line tool. Run the following command:

jvcli startproject my_project

This will create a new Jivas project named my_project in your current directory.

When you run the jvcli startproject command, the following files and folders are created:

actions/: This folder contains the action packages for your Jivas agents. Actions are the tasks that your agents can perform based on your specifications. Pre-built actions can be downloaded from the Jivas Package Repository (JPR) via the jvcli download command.
daf/: This folder is used for Digital Agent File (DAF) packages, which include all configurations, knowledge, and memory snapshots for your agents.
sh/: This folder contains optional shell scripts to help with initializing and managing your Jivas project.
- inituser.sh: Script to initialize a user and obtain a token.
- initagents.sh: Script to initialize agents.
- importagent.sh: Script to import a demo agent.
- serve.sh: Script to serve the Jivas app.
- startclient.sh: Script to launch the Jivas client.
tests/: This folder is for writing test cases for your project.
main.jac: The main entry point for your Jivas application, written in the Jac language.
.env: Environment configuration file for storing environment settings and secrets.
README.md: A guide to help you get started with your project.

Each of these components plays a crucial role in building and managing your Jivas AI project.

For more details, check out the rest of the documentation.