Skip to content

Installation & Setup

Install Prism

brew install scttfrdmn/tap/prism

scoop bucket add scttfrdmn https://github.com/scttfrdmn/scoop-bucket
scoop install prism

Download the latest tarball from the releases page, extract, and add to your PATH.

Verify: 

prism version

Connect AWS credentials

Prism uses your AWS credentials to launch instances in your account.

Step 1 — Configure the AWS CLI (skip if already done): 

aws configure
# Prompts for: AWS Access Key ID, Secret Access Key, region, output format

Verify it works: 

aws sts get-caller-identity

Step 2 — Add a Prism profile: 

prism profile add

This interactive wizard links a Prism profile name to your AWS credentials and default region. You only need to do this once.

For detailed IAM permission requirements, see the Administrator Guide.

First-time setup wizard

If this is your first time running Prism, the init wizard covers AWS setup, profile creation, and a test launch:

prism init

Launch your first workspace

# See what's available
prism templates

# Launch a Python + Jupyter environment
prism workspace launch python-ml my-first-project

# Check it's ready (takes ~2 minutes)
prism workspace list

# Get connection info
prism workspace connect my-first-project

prism workspace connect prints the SSH command and any web service URLs (Jupyter, RStudio, etc.).

Two interfaces

CLI

The prism command is the primary interface. It's scriptable and works in automated pipelines.

prism workspace launch python-ml my-project --size L --spot

GUI (desktop app)

A visual desktop app for managing workspaces, storage, and settings.

prism gui

Or launch Prism from your Applications folder (macOS) / Start menu (Windows).

Common workflows

Data science

prism workspace launch python-ml data-analysis --size L
prism workspace connect data-analysis
# Jupyter is available at http://localhost:8888 via SSH tunnel

R / statistics

prism workspace launch r-research stats-project
prism workspace connect stats-project
# RStudio Server available at http://localhost:8787 via SSH tunnel

Shared storage

prism volume create shared-datasets
prism volume attach shared-datasets my-project

Cost management

Stop workspaces when not in use — stopped instances have no compute cost:

prism workspace stop my-project
prism workspace start my-project   # Resume later

Hibernation preserves RAM state and reduces cost further:

prism workspace hibernate my-project
prism workspace resume my-project

Troubleshooting

"Daemon not running" 

prism admin daemon status
prism admin daemon stop     # Then run any prism command to auto-restart

"AWS credentials not found" 

aws sts get-caller-identity    # Verify credentials work
aws configure                  # Reconfigure if needed

Instance launch fails 

prism workspace launch python-ml my-project --region us-east-1

For more, see the Troubleshooting Guide.

Next steps