No description, website, or topics provided.
Clone or download
pitbulk Fix bug detected on the Protect MFA use. When PUSH Manual is used, ge…
…t_saml_assertion_verifying must contain do_not_notify=True
Latest commit 71bc386 Nov 2, 2018


Assume an AWS Role and get temporary credentials using Onelogin.

Users will be able to choose from among multiple AWS roles in multiple AWS accounts when they sign in using OneLogin in order to assume an AWS Role and obtain temporary AWS acccess credentials.

This is really useful for customers that run complex environments with multiple AWS accounts, roles and many different people that need periodic access as it saves manually generating and managing AWS credentials.

This repository contains a python script at src/onelogin/aws-assume-role/ that you can execute in order to retrieve the AWS credentials.

AWS and OneLogin prerequisites

The "Configuring SAML for Amazon Web Services (AWS) with Multiple Accounts and Roles" guide explains how to:

  • Add the AWS Multi Account app to OneLogin
  • Configure OneLogin as an Identity Provider for each AWS account
  • Add or update AWS Roles to use OneLogin as the SAML provider
  • Add external roles to give OneLogin access to your AWS accounts
  • Complete your AWS Multi Account configuration in OneLogin




The project is hosted at github. You can download it from:


The toolkit is hosted in pypi, you can find the python-saml package at


It works with python2 and python3.

  • boto3 AWS Python SDK
  • onelogin OneLogin Python SDK

Getting started


The use of a virtualenv is highly recommended.

Virtualenv helps isolating the python enviroment used to run the toolkit. You can find more details and an installation guide in the official documentation.

Once you have your virtualenv ready and loaded, then you can install the toolkit on it in development mode executing this:

python develop

Using this method of deployment the toolkit files will be linked instead of copied, so if you make changes on them you won't need to reinstall the toolkit.

If you want install it in a normal mode, execute:

python install


The python script uses a settings file, where OneLogin SDK properties are placed.

Is a json file named onelogin.sdk.json as follows

	"client_id": "",
	"client_secret": "",
	"region": "",
	"ip": ""


  • client_id Onelogin OAuth2 client ID
  • client_secret Onelogin OAuth2 client secret
  • region Indicates where the instance is hosted. Possible values: 'us' or 'eu'.
  • ip Indicates the IP to be used on the method to retrieve the SAMLResponse in order to bypass MFA if that IP was previously whitelisted.

For security reasons, IP only can be provided at the onelogin.sdk.json. On a shared machines where multiple users has access, That file should only be readable by the root of the machine that also controls the client_id / client_secret, and not by an end user, to prevent him manipulate the IP value.

Place that file in the same path where the python script is invoked.

Docker installation method

  • git clone
  • cd onelogin-python-aws-assume-role
  • Enter your credentials in the onelogin.sdk.json file as explained above
  • Save the onelogin.sdk.json file in the root directory of the repo
  • docker build . -t awsaccess:latest
  • docker run -it -v ~/.aws:/root/.aws awsaccess:latest python src/onelogin/aws-assume-role/ --onelogin-username {user_email} --onelogin-subdomain {subdomain} --onelogin-app-id {app_id} --aws-region {aws region} --profile default

How the process works

Step 1. Provide OneLogin data.

  • Provide OneLogin's username/mail and password to authenticate the user
  • Provide the OneLogin's App ID to identify the AWS app
  • Provide the domain of your OneLogin's instance.

Note: If you're bored typing your username (--onelogin-username), App ID (--onelogin-app-id), subdomain (--onelogin-subdomain) or AWS region (--aws-region) every time, you can specify these parameters as command-line arguments and the tool won't ask for them any more.

Note: Specifying your password directly with --onelogin-password is bad practice, you should use that flag together with password managers, eg. with the OSX Keychain: --onelogin-password $(security find-generic-password -a $USER -s onelogin -w), so your password won't be saved in you command line history. Please note that your password will be visible in your process list, if you use this flag (as the expanded command line arguments are part of the name of the process).

With that data, a SAMLResponse is retrieved. And possible AWS Role are retrieved.

Step 2. Select AWS Role to be assumed.

  • Provide the desired AWS Role to be assumed.
  • Provide the AWS Region instance (required in order to execute the AWS API call).

Step 3. AWS Credentials retrieved.

A temporal AWS AccessKey and secretKey are retrieved in addition to a sessionToken. Those data can be used to generate an AWS BasicSessionCredentials to be used in any AWS API SDK.

Quick Start using the python scrypt

Prepare the environment

After checking out the repository, let's use a virtual environment

cd <repository>
virtualenv venv

Activate then the environment

> source venv/bin/activate

Then run

> pip setup install


> develop

to install dependencies.


Assuming you have your AWS Multi Account app set up correctly and you’re using valid OneLogin API credentials stored on the onelogin.sdk.json placed at the root of the repository, using this tool is as simple as following the prompts.

> python src/onelogin/aws-assume-role/

Or alternately save them to your AWS credentials file to enable faster access from any terminal.

> python src/onelogin/aws-assume-role/ --profile profilename

By default, the credentials only last for 1 hour, but you can edit that restriction on AWS and set a max of 12h session duration.

Then set the -z or duration with the desired credentials session duration. The possible value to be used is limited by the AWS Role. Rea more at

You can also make it regenerate and update the credentials file by using the --loop option to specify the number of iterations, and --time to specify the minutes between iterations. If you specified a duration, be sure the value you set for the duration session of the credential is bigger than the value you set for the time, so your credentials will be renewed before expiration.

You can also make it interactive, with the -x or --interactiveoption, and at the end of the iteration, you will be asked if want to generate new credentials for a new user or a new role.

The selection of the AWS account and Role can be also be done with the --aws-account-id and --aws-role-name parameters.

For more info execute:

> python src/onelogin/aws-assume-role/ --help

Test your credentials with AWS CLI

AWS provide a CLI tool that makes remote access and management of resources super easy. If you don’t have it already then read more about it and install it from here.

For convenience you can simply copy and paste the temporary AWS access credentials generated above to set them as environment variables. This enables you to instantly use AWS CLI commands as the environment variables will take precedence over any credentials you may have in your ~/.aws directory.

Assuming that:

  • you have the AWS CLI installed
  • you have set the OneLogin generated temporary AWS credentials as environment variables
  • the role you selected has access to list EC2 instances

You should find success with the following AWS CLI command.

aws ec2 describe-instances


After checking out the repo, run pip setup install or python develop to install dependencies.

To release a new version, update the version number in src/onelogin/api/ and commit it, then you will be able to update it to pypy. with python sdist upload and python bdist_wheel upload. Create also a relase tag on github.


Bug reports and pull requests are welcome on GitHub at This project is intended to be a safe, welcoming space for collaboration, and contributors are expected to adhere to the Contributor Covenant code of conduct.


The gem is available as open source under the terms of the MIT License.

Code of Conduct

Everyone interacting in the OneLogin Assume AWS Role project’s codebases, issue trackers, chat rooms and mailing lists is expected to follow the code of conduct.