Quick Start

Note

The sc-machine can't be used on Windows.

Use sc-machine as a C++ library in your project

Conan

You can use Conan to install sc-machine. To integrate sc-machine into your project using Conan, follow these steps:

Create a conanfile.txt in your project root with the following content:
```
[requires]
sc-machine/<version>
```
Install pipx first using guide: https://pipx.pypa.io/stable/installation/.
Install Conan if not already installed:
```
pipx install conan
pipx ensurepath
```
Relaunch your shell after installation.
```
exec $SHELL
```

Add the OSTIS-AI remote Conan repository:

conan remote add ostis-ai https://conan.ostis.net/artifactory/api/conan/ostis-ai-sc-machine

Install sc-machine and its dependencies:
```
conan install . --build=missing
```
Import sc-machine targets into your CMake project by using:
```
find_package(sc-machine REQUIRED)
```
Start building! Refer to our C++ Guide on how to quickly develop an sc-machine agent in C++ from scratch.

GitHub Releases

You can download pre-built artifacts from GitHub Releases. Extract it to any location, then make it available to CMake by appending folder path to CMAKE_PREFIX_PATH:

set(FABULOUS_PROJECT_SC_MACHINE_PATH "/path/to/extracted/sc-machine" 
    CACHE PATH "sc-machine installation path"
)

list(APPEND CMAKE_PREFIX_PATH ${FABULOUS_PROJECT_SC_MACHINE_PATH})

Find more info on installation methods on the build system page (including some more advanced snippets for CMake).

Install sc-machine as a service

It's possible to use sc-machine through a WebSocket API. Docs for our WebSocket API clients in Python and TypeScript are available in the respective repositories. You might want to check sc-web for a more user-friendly interface as well.

Docker

We provide a Docker image for this project. Here's a quick snippet on how to launch it.

# clone the repo
git clone https://github.com/ostis-ai/sc-machine
cd sc-machine
# create empty knowledge base sources folder
mkdir kb
# note: at this stage you can move your KB sources to the ./kb folder

# build knowledge base
docker compose run --rm machine build
# run sc-machine
docker compose up

Head to Installing with Docker to learn more. Using Docker is the recommended way to deploy sc-machine in production.

Native

Note: currently, running sc-machine natively on Windows isn't supported.

Download the sc-machine release for your platform and run it on your system:

# create empty knowledge base sources folder
mkdir kb
# note: at this stage you can move your KB sources to the ./kb folder

cd sc-machine-<version>-<platform>
# build knowledge base
./bin/sc-builder -i ../kb -o ../kb.bin --clear
# run sc-machine
./bin/sc-machine -e lib/extensions -s ../kb.bin 
# kb.bin will become your KB persistence folder

In case you want to make changes to the project sources, please refer to the build system docs.

Config

This repository provides a default configuration for the sc-machine. To customize the sc-machine to suit your needs you can create your own config file.