- These variables have been moved to a config file to be inserted into the task definition on deploy
- Modify the docker-start script to turn the comma separated key-value pairs in the ENV_VARS secret into environment variables

- Limits permission scope for production deployments
- Adheres to the principle of least privilege
- Separates permissions between application deployments and infrastructure changes

- The application can now be deployed independently from infrastructure changes
  - Create SSM parameter population script in python so it can also be used between services located in different repositories (e.g. like reporting service)
- Create a new task definition based on the templated version created by infrastructure deploy
- Create appspec.yaml template for handling codedeploys
- Remove redundant parameters from terraform deployment

- This is required due to the new template-mechanism
- Adheres to split between application and infrastructure

- Remnant from database migration

- Dashes instead of underline for github pipelines
- Clear separation of concerns in account tf stack
- Remove unnecessary `app_version` input

- Docker entry point is top level so this is where the environment
  variables must be extracted
- Otherwise shelling into the container and also db:seed does not
  execute properly

Also format some yml whitespace using prettier.

Matches the one in .tool-versions.

The workflow always replaces the db cluster, so the check isn't
necessary anymore.

Update the title, pass in `git_ref_to_deploy`, only
`build-and-push-image` image if an `image_tag` isn't specified, and
ensure that `yq` is installed before using it.

Skip building if an `image-tag` is passed in, and rename the
`deploy-sidekiq` subtask for consistency.

Use kebab-case for variables and remove an unused env var.

We don't deploy this worker anymore.

- Agreed to change the setup to control the set of tunable variables
  only on infrastructure side
- Parameter store locations for the variables are created, but only
  populated with a CHANGE_ME parameter
- This is for use with non-prod systems only
- The docker entry point now unsets any variables with CHANGE_ME values
- Workflow updated to reflect new secret/aws variable management setup
- python scripts/dependency removed as no scripting is needed for
  secrets in the new setup

Currently these are not set in config and must therefore be included
here

- Fixes SonarQube security warnings

- Aligns deployment flows accross repositories
- Ensures uniqueness and covers edge-cases for non-existance of task
  definitions