Plugin – lektor-s3 0.5.1

Publish to S3 buckets and Cloudfront.

Project links

GitHub Statistics


Version: 0.5.1

Author: Spencer Nelson


AWS, CloudFront, publisher, S3, and setup-env

View all tags.

Project Description


Build Status

lektor-s3 makes it easy to deploy your Lektor project to an S3 bucket.

Before you start

You're going to be storing your website's data in an S3 bucket. The code here won't do anything to create or configure that bucket. You'll have to create the S3 bucket and set it up yourself.

AWS has a pretty good guide for how to set up a bucket to host a static website. You'll need to both create the bucket and set its permissions to allow global read access. Remember to do this first because lektor-s3 won't do it automatically.

Installation and Usage

Install with the usual Lektor toolchain. Within your project, run

lektor plugins add lektor-s3

You should see a message saying lektor-s3 has been added to the project.

Next, add an S3 bucket to your project's servers. In your project file (like blog.lektorproject), add the following:

name = S3
enabled = yes
target = s3://<YOUR-BUCKET>

For example, if you wanted to deploy to a bucket named 'huntedwumpus', you'd make that last line

target = s3://huntedwumpus

Now, if you call lektor deploy s3, Lektor will upload your built website to S3 in the bucket you targeted.


Optionally, you can also provide a CloudFront distribution ID. If you do, Lektor will invalidate all objects in that CloudFront distribution after every deploy.



You need to prove to S3 that you have permission to upload to the bucket you've chosen.

lektor-s3 uses boto, which means it will obey boto's usual flow for gathering credentials.

For a refresher, that means you have two options: you can store your credentials in an INI file at ~/.aws/credentials, or you can pass credentials in through the environment variables AWS_ACCESS_KEY_ID and AWS_SECRET_ACCESS_KEY.

If you have multiple sets of credentials, you can group them into profiles in your credentials file and choose the right one using the AWS_PROFILE environment variable. Your ~/.aws/credenials file might look like this:

aws_access_key_id = <...>
aws_secret_access_key = <...>

aws_access_key_id = <...>
aws_secret_access_key = <...>

And then you can invoke lektor with the environment variable:

$ AWS_PROFILE=personal lektor deploy`


You can specify headers to be attached to particular files when uploading them to S3. These can be configured in an INI file at configs/s3.ini under your project root.

You can name the sections anything that makes sense to you, but every section must have either a match or an extensions item to specify which files the configuration applies to. If using match, you should write this as a regular expression that will be applied against the filename using the regular expression's search method. If using extensions, write a comma-separated list of the file extensions to which the configuration applies. Both match and extensions may be specified.

The rest of the items in each section should specify one or more headers and their values. A list of valid headers is defined in the boto documentation as ALLOWED_UPLOAD_ARGS.

Defaults can be defined via the usual INI file way, in a [DEFAULTS] section.

For example, your configuration file might look like this:

CacheControl = public,max-age=3600

[static files]
match = \.(css|js|woff|woff2)$
CacheControl = public,max-age=31536000

extensions = jpg,jpeg,png,mp4
CacheControl = public,max-age=259200

extensions = woff2
ContentType = application/font-woff2

extensions = html,txt
ContentLanguage = en


Pull requests are super useful and encouraged! Once accepted, changes are published using lektor with lektor dev publish-plugin.

Run your tests by invoking python test.