Namespace Kuestenlogik.Bowire.Flows

FlowDataGenerator: Row generator config: range (from..to inclusive) or random (seeded, reproducible).

FlowDataRow: One expanded data row — the column values join the {{var}} resolver scope for a single step execution, the label names the row in reports (step[label]).

FlowDataSource: Data-driven parameterisation for a Flow step — #174. Exactly one of Inline / Csv / Generator is set; the runner expands the step into one execution per row, with the row's columns joining the {{var}} resolver scope (row wins over --env).

FlowDataSourceExpander: Materialises a FlowDataSource into concrete rows — #174. The runner calls this once per data-driven step and executes the step once per returned row.

FlowDefinition: Server-side projection of the workbench's Flow document — the v2.1 in-browser editor persists the canonical shape into localStorage[bowire_flows]; this record mirrors only the fields the v2.2 CLI runner (T2) and the C# expectation evaluator need. Streaming / Foreach / Variable / Condition / Loop nodes are represented but their bodies stay opaque (raw JsonNode) because T2 only EVALUATES — replaying them is the CLI runner's job.

FlowStep: One step inside a Flow. Each step type uses different fields; common ones are first, type-specific ones tail off into Body / Service / etc. Lives separately from CapturedFlow (Proxy capture record) — the names collide intentionally, the domains don't.

LegacyAssertionTuple: Legacy v2.1 assertion tuple — the in-browser editor wrote {path, op, value} before the v2.2 schema landed. Kept on the step so a saved flow round-trips cleanly through the new pipeline without forcing a one-shot migration.

Sections