If you're on e.g. CoreOS and you don't want to have to install Python or other libraries, use Docker to access S3.
See [http://docs.aws.amazon.com/cli/latest/userguide/cli-chap-getting-set-up.html](Amazon's help docs) on how to get your access key and credentials
After you have those, run this command to save your AWS creds locally (to the folder /my/current/directory/.awscreds)
docker run --rm -t -i -v $(pwd)/.awscreds:/root/.aws alexturek/aws-cli-docker configure
Enter your key ID and secret (and any other information) at the prompts. These are now saved at ./creds/
From the same directory:
docker run --rm -t -i -v $(pwd)/.awscreds:/root/.aws alexturek/aws-cli-docker my aws commands
If you want to be able to run this from any directory on your machine, replace $(pwd) in both commands with /some/absolute/path
See objects in an S3 bucket:
docker run --rm -t -i -v $(pwd)/.awscreds:/root/.aws alexturek/aws-cli-docker s3 ls s3://my-s3-bucket
For more use see [http://docs.aws.amazon.com/cli/latest/reference/](AWS's documentation)