AWS Lambda Tag a Function for Cost Allocation

arazzo: 1.0.1
info:
  title: AWS Lambda Tag a Function for Cost Allocation
  summary: Resolve a function name to its ARN, apply cost-allocation and ownership tags, and read the tags back to confirm the merge.
  description: >-
    Tagging is how Lambda spend is attributed to a team, environment, or cost
    center in AWS Cost Explorer, and the tagging calls are the awkward corner of
    this API: they are keyed by the function's full ARN rather than its name, so
    a plain function name cannot be tagged directly. This workflow resolves the
    name to an ARN first, applies the tags, and reads them back. Tagging merges
    rather than replaces, so keys not mentioned survive untouched — which is why
    the read-back matters when several pipelines tag the same function. Every
    step spells out its request inline so the flow can be read and executed
    without opening the underlying OpenAPI description.
  version: 1.0.0
sourceDescriptions:
- name: lambdaApi
  url: ../openapi/aws-lambda-api-openapi.yml
  type: openapi
workflows:
- workflowId: tag-function
  summary: Resolve a function's ARN, apply tags to it, and verify the resulting tag set.
  description: >-
    Looks up the function to obtain the ARN the tagging endpoints require,
    applies a set of key-value tags, and lists the tags afterwards to confirm the
    merged result.
  inputs:
    type: object
    required:
    - functionName
    - tags
    properties:
      functionName:
        type: string
        description: The name or partial ARN of the function to tag.
      tags:
        type: object
        description: >-
          Key-value pairs to apply (e.g. {"Environment":"production",
          "CostCenter":"platform","Owner":"payments-team"}). Existing keys not
          listed here are left in place.
  steps:
  - stepId: resolveFunctionArn
    description: >-
      Look up the function to obtain its full ARN. The tag endpoints are keyed by
      ARN, not by function name, so this lookup is a hard requirement rather than
      a convenience. The response also carries the tags already on the function.
    operationId: getFunction
    parameters:
    - name: FunctionName
      in: path
      value: $inputs.functionName
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      functionArn: $response.body#/Configuration/FunctionArn
      existingTags: $response.body#/Tags
  - stepId: applyTags
    description: >-
      Apply the tags to the resolved ARN. The call merges: keys present are set
      or overwritten, and keys absent from this payload are left alone. It
      returns 204 with no body, which is why the tag set has to be read back to
      be observed.
    operationId: tagResource
    parameters:
    - name: ARN
      in: path
      value: $steps.resolveFunctionArn.outputs.functionArn
    requestBody:
      contentType: application/json
      payload:
        Tags: $inputs.tags
    successCriteria:
    - condition: $statusCode == 204
  - stepId: verifyTags
    description: >-
      List the tags on the function to confirm the merged result, showing both
      the newly applied keys and any pre-existing keys that survived.
    operationId: listTags
    parameters:
    - name: ARN
      in: path
      value: $steps.resolveFunctionArn.outputs.functionArn
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      tags: $response.body#/Tags
  outputs:
    functionArn: $steps.resolveFunctionArn.outputs.functionArn
    previousTags: $steps.resolveFunctionArn.outputs.existingTags
    tags: $steps.verifyTags.outputs.tags