Clear Workflow Context

The Clear Workflow Context job removes selected data from the current workflow context. Use it to reset temporary state between workflow steps, clean up after a branch, or make sure downstream jobs do not accidentally reuse earlier metadata or resources.

Properties

The properties for configuring the cleanup are described below.

• Clear temporary group (optional; default: false)
  • Removes all members from the workflow context temporary group.
• Clear groups (optional; default: false)
  • Removes workflow groups from the workflow context.
• Clear incoming unit (optional; default: false)
  • Removes the IncomingUnit resource from the workflow context.
• Clear files (optional; default: false)
  • Removes files stored in the workflow context.
• Clear incoming group member (optional; default: false)
  • Removes the IncomingGroupMember resource from the workflow context.
• Clear meta data (optional)
  • Regular expression used to match metadata keys that should be removed.
• Clear resources (optional)
  • Regular expression used to match resource names that should be removed.

Referencing syntax

The two pattern-based properties use regular expressions.

• Leave Clear meta data empty to keep all metadata.
• Use .* in Clear meta data to remove all metadata keys.
• Leave Clear resources empty to keep all resources.
• Use targeted patterns such as ^tmp_ when you want to clear only temporary keys or resources with a naming convention.
• Use the format ^name$ when you want to match one exact metadata key or one exact resource name only. For example, ^customer_id$ matches only that metadata key, and ^request$ matches only a resource named request.

Cleanup behavior

Each option affects a different part of the workflow context.

• Clearing groups removes fallback groups that some downstream jobs rely on.
• Clearing the incoming unit or incoming group member can cause later jobs to fail if they require those resources.
• Regex-based clearing is applied by name, so the pattern should be reviewed carefully before using broad matches.

Best practices and tips

• Clear only the parts of the workflow context you no longer need.
• Prefer targeted regex patterns over .* unless a full reset is intentional.
• Adopt temporary naming conventions such as tmp_ or another clear prefix so cleanup patterns stay precise and safe.
• Clean up after a branch or subprocess has finished, not before. Downstream jobs often depend on incoming units, groups, or response resources remaining available until the end of that path.
• When a workflow later depends on IncomingUnit, groups, or files, clear those values only after the dependent jobs have finished.

Common cleanup patterns

Remove only temporary resources

If temporary resources follow a naming convention, clear them by prefix:

• Clear resources: ^tmp_

Clear specific resources after an iteration

To remove only known intermediate resources, match exact names:

• Clear resources: ^call_log_response$|^call_log_json$

Targeted metadata cleanup

To remove one metadata key only:

• Clear meta data: ^retry_count$

Targeted cleanup keeps required context intact while preventing stale temporary values from leaking into later branches.

Related jobs

• Filter Workflow Context
• Split Workflow Context
• Workflow Context

References

• Working With Variables
  • Background on metadata and resource naming within the workflow context.
• Working With Strings
  • Useful when writing regex patterns for metadata or resource cleanup.