upload-to-s3-with-taskcluster.py

import taskcluster, boto.s3

auth = taskcluster.sync.Auth()
result = auth.awsS3Credentials('read-write', 'tc-gp-public-31d', 'garbage/')

s3 = boto.s3.connect_to_region(
  'us-west-2',
  aws_access_key_id=result['credentials']['accessKeyId'],
  aws_secret_access_key=result['credentials']['secretAccessKey'],
  security_token=result['credentials']['sessionToken'],
)

bucket = s3.get_bucket('tc-gp-public-31d', validate=False)
k = bucket.new_key('garbage/test.txt')
k.set_contents_from_string("Hello World")

print "https://tc-gp-public-31d.s3-us-west-2.amazonaws.com/garbage/test.txt"

pip install taskcluster boto first...

But really we should probably be switching to boto3.