Identify the resource
Give Kato the name and kind of resource, its workspace if relevant, and an identifier when available. For example:Find the ingestion job for [collection name] in [workspace name]. Show me the job and its latest run, including their identifiers. Do not start another run.If names are ambiguous, supply the intended dataset, owner or run identifier. For time-based work, state the exact period and time zone. Use dates and a time zone appropriate to your own inputs when the task is time-based. Ask Kato to identify the match before making changes. No match means no resource was found within the permitted search; it does not establish that none exists. Check the name and scope with the resource Owner if needed.
Ask about the work record
For collection, ask what was accepted, failed or skipped and which processing remains incomplete. For an evaluation, ask which candidate and inputs were tested and which cases failed. For a workflow, ask for the selected definition version, current attempts, outputs and human waits.Inspect this run. What has completed, what failed and what is waiting for a person or still processing? Give me links to the records I need to check. Do not retry, cancel or approve anything.Kato uses your current read access. It may be unable to show a record, output or status detail that you cannot read.
Open and verify the result
Ask Kato to open the identified resource. Check its title, breadcrumb trail and identifier, then inspect the linked work record. Kato can navigate to supported resource pages such as a workspace, dataset, workflow or workflow run; some related records are reached from their parent page.
A completed run does not establish that every source was available, a conclusion was supported or an output was approved and published. Inspect the appropriate evidence and decision records.