Google Drive Back Up a File Before Permanently Deleting It

arazzo: 1.0.1
info:
  title: Google Drive Back Up a File Before Permanently Deleting It
  summary: Resolve a file, copy it into an archive folder, verify the backup exists, and only then permanently delete the original.
  description: >-
    Drive's delete endpoint permanently deletes the file without moving it to
    the trash, which means there is no undo and no 30-day recovery window — the
    usual safety net simply is not there. Any automated cleanup job therefore
    has to create its own. This workflow resolves the target file, copies it
    into an archive folder, reads that copy back to prove the backup actually
    exists, and only then issues the irreversible delete. The ordering is the
    whole point: the destructive step runs last and only after the backup has
    been independently verified. 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: filesApi
  url: ../openapi/google-drive-files-api-openapi.yml
  type: openapi
workflows:
- workflowId: backup-and-delete-file
  summary: Archive a copy of a file, verify it, then permanently delete the original.
  description: >-
    Reads the file's metadata, copies it into an archive folder under a
    timestamped-style archive name, confirms the archived copy is readable, and
    then permanently deletes the original file.
  inputs:
    type: object
    required:
    - fileId
    - archiveFolderId
    - archiveName
    properties:
      fileId:
        type: string
        description: The ID of the file to archive and then permanently delete.
      archiveFolderId:
        type: string
        description: The ID of the folder the backup copy should be created in.
      archiveName:
        type: string
        description: >-
          The name to give the archived copy (e.g. "2026-07-17 - Q3 Report
          (archived)").
  steps:
  - stepId: resolveFile
    description: >-
      Read the file's metadata to confirm it exists and capture its name and
      size before anything destructive is attempted.
    operationId: getFile
    parameters:
    - name: fileId
      in: path
      value: $inputs.fileId
    - name: fields
      in: query
      value: id,name,mimeType,size,parents,modifiedTime,trashed
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      resolvedFileId: $response.body#/id
      originalName: $response.body#/name
      originalMimeType: $response.body#/mimeType
      originalSize: $response.body#/size
  - stepId: archiveCopy
    description: >-
      Copy the file into the archive folder under the supplied archive name.
      This copy is the only recoverable artifact once the original is deleted.
    operationId: copyFile
    parameters:
    - name: fileId
      in: path
      value: $steps.resolveFile.outputs.resolvedFileId
    requestBody:
      contentType: application/json
      payload:
        name: $inputs.archiveName
        parents:
        - $inputs.archiveFolderId
    successCriteria:
    - condition: $statusCode == 200
    outputs:
      archiveFileId: $response.body#/id
  - stepId: verifyBackup
    description: >-
      Read the archived copy back and confirm it is present, untrashed, and
      filed in the archive folder. This step is the gate on the delete — if the
      backup cannot be read here, the original must not be destroyed.
    operationId: getFile
    parameters:
    - name: fileId
      in: path
      value: $steps.archiveCopy.outputs.archiveFileId
    - name: fields
      in: query
      value: id,name,mimeType,size,parents,trashed,webViewLink
    successCriteria:
    - condition: $statusCode == 200
    - context: $response.body
      condition: $.trashed == false
      type: jsonpath
    outputs:
      verifiedArchiveId: $response.body#/id
      verifiedArchiveName: $response.body#/name
      archiveWebViewLink: $response.body#/webViewLink
  - stepId: deleteOriginal
    description: >-
      Permanently delete the original file. This bypasses the trash and cannot
      be undone, which is why it runs only after the archived copy has been read
      back successfully.
    operationId: deleteFile
    parameters:
    - name: fileId
      in: path
      value: $steps.resolveFile.outputs.resolvedFileId
    successCriteria:
    - condition: $statusCode == 204
    outputs:
      deletedFileId: $steps.resolveFile.outputs.resolvedFileId
  outputs:
    originalName: $steps.resolveFile.outputs.originalName
    deletedFileId: $steps.deleteOriginal.outputs.deletedFileId
    archiveFileId: $steps.verifyBackup.outputs.verifiedArchiveId
    archiveWebViewLink: $steps.verifyBackup.outputs.archiveWebViewLink