Operations And Timeline
An operation is the review trail for multi-step bot or runtime work. It shows what was requested, what the agent did, which tools were called, what outputs were produced, and how the work ended.
What The Timeline Shows
- User or channel input that started the work.
- Agent turns and reasoning summaries visible to the product.
- Tool calls, callbacks, outputs, and errors.
- Artifacts, links, events, or messages created by the work.
- Final status such as completed, failed, canceled, or waiting for action.
How Users Should Read It
Start at the request, then follow each tool step and output. If an operation produced an artifact, open that artifact and inspect its source or rendered blob. If it failed, capture the failing step and visible error before retrying.
Why It Matters
Operations make automated work accountable. A user can see whether a result came from source inspection, a dataset query, a forecast, a rendered file, or an unsupported assumption.
Media To Add
- Screenshot: complete operation timeline with request, tool call, output artifact, and final status labeled. It helps users review agent work. Source: sample bot operation.
- Video: failed operation review and retry path. It helps support and admins understand recovery. Source: test workspace with a harmless failing tool call.