A configuration is one exact setup: coding tool, tool version, model, and reasoning effort. Change any of them and it becomes a new configuration.
Grouped by tool for browsing only. Counts show how much is archived, not how good anything is.
The launch cohort: model and reasoning settings could be the tool’s defaults, and subagents were not ruled out. Runs are compared only within one protocol. How the protocols differ
Claude Opus 5 (1M context) on Claude Code
Operator-attested: the Configuration, tool version, displayed model, and elapsed time were recorded by the operator from what the tool showed at Run time. They are not independently verified. An exact model ID appears only when the tool exposed one; it is never inferred.