Before you can configure pxl, you need a few things:

  • A pxl installation, see the installation guide here.
  • Credentials to an S3 bucket. You need a Cloud Provider to give these to you.
  • Credentials for deployment. (An SSH user, and a path to upload the generated site).

If you've got this; you're ready to move!

Config file

pxl reads config from ~/.config/pxl/config.json. This is a JSON file with a top level object from which pxl understands the following config keys:

  • "s3_endpoint"
  • "s3_region"
  • "s3_bucket"
  • "s3_key_id"
  • "s3_key_secret"
  • "deploy_host"
  • "deploy_user"
  • "deploy_path"
  • "public_image_url"

You can write this file yourself, or you can use the setup wizard below. In case pxl ever gets new settings, it is probably good to know that this file exists.

This is an example config file:

  "s3_endpoint": "",
  "s3_region": "ams3",
  "s3_bucket": "pxl-demo",
  "deploy_host": "",
  "deploy_user": "pxl",
  "deploy_path": "",
  "public_image_url": ""

Setup wizard

This is a small convenience to get going with pxl. If you don't want to write the config file by hand, you can let pxl generate it.

In your copy of the pxl repository, you can run the pxl init command. If you answer the questions like below (just hitting the enter key uses the default), you'll get the same config file as we showed you before.

$ pipenv run pxl init
We need some information. Please answer the prompts.
Defaults are between [brackets].

S3 endpoint []:
S3 region [ams3]:
S3 bucket: pxl-demo
S3 key secret (not shown):
Deploy host:
Deploy user: pxl
Deploy path:
Public image base URL (optional) []: