Use Case: Mitigate stuck jobs when no worker can pick them up

[grafvonb]

Sometimes a Camunda process instance waits because a job was created, but no worker will ever pick it up.

A typical example is a service task with a wrong job type. The process instance looks stuck, but the actual reason is not obvious enough for an operator.

c8volt should support an operational workflow for this situation:

c8volt ops mitigate stuck-jobs

The workflow should help operators find jobs that look stuck, preview the affected jobs, and then explicitly fail them with a meaningful diagnostic message.

Example:

c8volt ops mitigate stuck-jobs \
  --type payment-worker \
  --older-than 30m \
  --message "no worker available for job type payment-worker" \
  --dry-run

This is not intended to repair the process automatically. It is intended to make a silent operational problem visible and actionable by creating explicit job failures or incidents.

Expected behavior:

• show a dry-run before changing anything
• require an age filter so fresh jobs are not touched accidentally
• require a diagnostic message from the operator
• require confirmation before mutation
• provide machine-readable output for automation
• optionally write an audit report

Implementation details are tracked in:

#232