Use Splitrail when you need a live operational view of token burn, cost, and model mix across multiple coding-agent tools during or after runs. Invoke it instead of checking each product’s native dashboard when the job is cross-tool usage monitoring and cost triage, not everyday agent use. The scope boundary is spend observability for local agent runs, not a general coding assistant or cloud platform listing.