This PR adds quite a few bits of new functionality and refactor (some of which might be a step too far for you, but I thought I'd give it a go anyway :)

• chart renamed to vsts-agent instead of helm-vsts-agent
• make helm-package task added to create package (can now use helm repo add vsts-agent https://raw.githubusercontent.com/grenzr/helm-vsts-agent/assorted_improvements/charts/ to pull in the repo)
• add extraEnv variable to add more env vars into the vsts-agent container
• add cleanRun which resets vsts-agent container upon completion of each build (resets the working environment)
• add extraContainers variable which allows sidecar of additional containers into the vsts-agent pod
• add volumes/volumeMounts variable which allows ability to mount arbitrary volumes into the vsts-agent container or elesewhere in the pod

Our example use is to sidecar our own docker DIND image instead of using the k8s hosts docker service (which imo is not a good approach). This way guarantees consistency from k8s deployment to deployment :

resources:
  limits:
    cpu: 1
    memory: 1Gi
  requests:
    cpu: 1
    memory: 1Gi

cleanRun: true

extraEnv:
  DOCKER_HOST: tcp://localhost:2375

extraContainers:
- name: dind-daemon
  image: docker:18.05.0-dind
  args: ["--mtu=1450"]
  resources:
    requests:
      memory: 1Gi
  securityContext:
      privileged: true
  volumeMounts:
    - name: docker-graph-storage
      mountPath: /var/lib/docker

volumes:
- name: docker-graph-storage
  emptyDir: {}

volumeMounts: []