Skip to content


Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time

Azure Backlog Generator (AzBacklog)

github codecov PyPI version

Following repetitive processes and aligning all resources consistently can be a daunting task, and it can be common to overlook a step or make an incorrect assumption. There are many variables to consider in any complex process—identifying roles, resources, etc.

When conducting intense processes, it can be helpful to have a list of tasks to direct efforts. The Azure Backlog Generator (AzBacklog) is designed to build backlogs for complex processes based on proven practices. The backlogs can be generated in either Azure DevOps or GitHub.

Currently Supported Backlogs

AzBacklog currently supports creating backlogs for the following processes.

Backlog Command-line Option (see below: -b)
Cloud Adoption Framework (CAF) caf
Enterprise-scale Architecture esa
(Coming Soon) Team Foundation Server (TFS) to Azure DevOps tfs


Prior to execution, please take a moment to fully read through the documentation below.


First, install the azbacklog package:

pip install azbacklog

NOTE: The library installs additional data files (the workitems) into a artifacts folder that your user may not have access to under the normal user permissions. MAC users using Homebrew may especially have this issue. In this case, instead of using sudo to install the package, use the --user to install the package into the localized user's libary (ex: pip install azbacklog --user). Finally, by installing to the localized user's path, you may need to add the path's binary folder to Python's search path.

Create a Backlog

To create a backlog:

./azbacklog -t TOKEN

usage: azbacklog [-h] -t TOKEN [-r {azure,github}] [-p PROJECT] [-o ORG] [-b {caf,tfs}]

Generate a backlog of work items.

required arguments:
  -t TOKEN, --token TOKEN                           GitHub or Azure DevOps token

optional arguments:
  -h, --help                                        show this help message and exit
  -r {azure,github}, --repo {azure,github}          targetted repository type
  -p PROJECT, --project PROJECT                     project (repository) name to create
  -o ORG, --org ORG                                 REQUIRED. if target is Azure DevOps
                                                    OPTIONAL. if a GitHub organization
  -b {caf,esa}, --backlog {caf,esa}                 type of backlog to create


To contribute to the project from your development environment:

git clone generator
cd generator

pip install -e .[dev]

The previous commands will clone the repo and not only install the runtime dependencies but also the development dependencies.

Executing Locally Under Development

Once you've installed the package locally (e.g. -e .[dev] in the previous command), executing the package is as simple as (include the necessary parameters outlined above in the Create a Backlog section):


Running Tests

To simply execute unit tests:

python -m pytest

If you would like to view a coverage report of the unit tests:

# Generate the report using pytest
coverage run -m pytest

# View the report, including missing lines
coverage report -m

Shared Responsibility

Microsoft maintains a position of shared responsibility between itself and the customer. Under this shared responsibility, it is understood that appropriate cloud adoption requires active participation from the architect(s) involved and the customer unit(s). In this process, the customer should be prepared to make available network and identity administrators, application and data architects, and other security principals. By understanding requirements through active feedback, Microsoft and its technical architects can adequately assess customer needs and propose recommendations that are proven in practice and that successfully meet customer objectives.

Technical Requirements

The scripts in this repository require Python as they rely on native Azure DevOps and GitHub SDKs. To execute these scripts, please download Python.

Technical Guidelines

Below are the technical guidelines that should be followed when contributing to the project.


The application, including unit tests, are developed in Python in order to maintain compatibility between Azure DevOps and GitHub SDKs. The backlog items are developed in simple JSON format with very little metadata.

Backlog Items

The backlog items are arranged in a collection of epics, features, stories and tasks with the necessary parent-child relationships. The following table outlines the correlations between the backlog items in Azure DevOps and GitHub.

Azure DevOps GitHub Issues
Epic Project
Feature Milestone
User Story Issue
Task Checklist in the body of the Issue

Excluding the specific git action of 'tagging' a commit, tags have differing nomenclature between Azure DevOps and GitHub. In Azure DevOps, tags are associated with work items whereas, in GitHub, the terminology for tags is "labels." That being the case, and based on the previous table, tags will be associated with all work item types in Azure DevOps, but will only be associated with issues in GitHub.

AzBacklog will create tags, as applicable, in both platforms--Azure DevOps and GitHub.

Creating Custom Backlogs

Contributors are welcome to create new backlogs that others may find useful for repititive actions. The following sections will explain the process.

Directory Structure

The directory structure follows a very simplistic heirarchy to fascilitate dependency creation and increase efficiency by minimizing overhead:

├── src                         (source code)
├── tests                       (unit tests)
└── workitems
    └── backlog_type
        ├── config.json         (specific backlog configuration)
        ├──           (optional. for repo initialization)
        ├── 01_SampleEpic1
        │   ├── metadata.json
        │   └── 01_SampleFeature1
        │       ├── metadata.json
        │       └── 01_SampleStory1
        │           ├── metadata.json
        │           ├── 01_SampleTask1
        │            │   └── metadata.json
        │            └── 02_SampleTask2
        │                └── metadata.json
        └── 02_SampleEpic2
            ├── metadata.json
            ├── 01_SampleFeature1
            │   ├── metadata.json
            │   └── 01_SampleStory1
            │       ├── metadata.json
            │       └── 01_SampleTask1
            │           └── metadata.json
            └── 02_SampleFeature2
                ├── metadata.json
                └── 01_SampleStory1
                    ├── metadata.json
                    ├── 01_SampleTask1
                    │   └── metadata.json
                    └── 02_SampleTask2
                        └── metadata.json

The heirarchy above follows the structure of Epic -> Feature -> User Story -> Task. Each backlog type must contain a config.json that directs the engine's creation process. Each folder will contain a metadata.json file that follows the Work Item Format below and child items. Task folders will not contain any child work items and should only contain its metadata. Finally, the top folder for each backlog type may contain a that will be used upon repository initialization.

Config.json Format

The format of the backlog's config.json file is the following:

    "description": "My backlog",
    "tags" : [
        "First Workload",
        "First Host",
        "Workload Template"
    "roles": [
    "tagcolors": [


A simple, single-line description of the backlog.

Allowed tags that may be used for the work items. Tags can be used to group similar work item types.

Allowed roles that may be used for the work items. Roles can be used to easily identify team individuals who should be assigned to the work item.

Hexidecimal colors to be used for each tag and role. Standard colors can be used for easier identification based on business process.

NOTES: 1) The total number of tagcolors must equal the combined sum of the tags and roles. 2) tagcolors are not required when deploying to Azure DevOps, but are required for GitHub deployments.

Work Item Format

The format of a work item's metadata.json file is the following:

    "title": "Work Item Title",
    "description": "Some description of the work item",
    "tag": ["Strategy | Plan | Ready"],
    "roles": ["AppDev"]


A simple, single-line title of the work item.

The content of the work item.

The tag(s) for the specific work item.

The role(s) for the specific work item.

NOTE: Tags and Roles must be those provided in the respective list of the corresponding process. The available lists are found with the backlog descriptions.


Your experience and feedback are valuable and, therefore, your contributions are welcomed. Please create necessary issues and, optionally, pull requests for your feedback or contributions. Please adhere to the technical guidelines above when contributing to the source code.

Most contributions require you to agree to a Contributor License Agreement (CLA) declaring that you have the right to, and actually do, grant us the rights to use your contribution. For details, visit

When you submit a pull request, a CLA bot will automatically determine whether you need to provide a CLA and decorate the PR appropriately (e.g., status check, comment). Simply follow the instructions provided by the bot. You will only need to do this once across all repos using our CLA.

This project has adopted the Microsoft Open Source Code of Conduct. For more information see the Code of Conduct FAQ or contact with any additional questions or comments.


Create a backlog of actionable work items in alignment with various processes based on proven practices.



Code of conduct





No packages published