Passing data into workflow steps

Use parameters and outputs to pass data into your workflow steps.

Use a parameter to pass external data into a workflow step, and an output to pass data from one step into another step in a workflow.

Passing external data into a step

To pass external data into a workflow step, first declare it in the top-level parameters: map, then use the !Parameter type in your step specification.

This Slack notification example uses parameters to pass a Slack channel name and a custom message into the workflow step, which uses a curated container image. When you run the workflow, Relay asks you to enter values for channel and message.

parameters:
  channel:
    description: Slack channel (include preceding hashtag)
    default: "#relay-workflows"
  message:
    description: Message to send to the channel
steps:
  - name: notify-channel
    image: projectnebula/slack-notification
    spec:
      channel: !Parameter channel
      apitoken: !Secret slack-token

For details about the parameters map, see the Workflow reference

Passing internal workflow data into a step

If you need to pass data from one step in your workflow to another step, use the !Output type in the spec section of the step which consumes the data. The step which produces the data should use the ni output set command to do so, as in this example:

parameters:
  message:
    description: "The message to output from the final step"
steps:
- name: generated-output
  image: projectnebula/core
  input:
  - ni output set --key dynamic --value "$(/bin/date)"
- name: hello-world
  image: projectnebula/core
  spec:
    message: !Parameter message
    dynamic: !Output [generated-output,dynamic]
  input:
  - echo "Hello world. Your message was $(ni get -p {.message}), and the generated output was $(ni get -p {.dynamic})"

Advanced users can use the Python SDK in scripts to create and modify keys instead of running ni.