Reading run history and expanding a run to per-step outputs
Every time a flow runs, HIIE records it so you can review what happened. The Runs table is where you read that history.
Where run history comes from
Each completed run is persisted as a run record (a run_*.json file) in the project's Flows folder. The Runs table reads these records and displays them as your run history.
How the table is ordered
Runs are shown newest-first, with any currently running flows pinned to the top. That makes it easy to keep an eye on in-progress work while still having the full past history below it.
Expanding a run
Each run in the table is expandable. Open one and you'll see its per-step outputs — what each step in the DAG produced — along with the artifacts that run saved into your project. Because each step logs an Agent-to-Agent envelope, the whole run also reads as a single conversation in the node graphs.
What the per-step view tells you
The per-step breakdown is the best way to understand and debug a flow. You can see which role or integration ran at each step, what it returned, and which output was carried forward to the steps that depended on it. If a step saved an artifact, you'll see it labeled by type — Research, Analysis, Design, Spec, Legal, Manufacturing for role steps, or Integration for MCP steps.
Using history effectively
Reviewing past runs helps you refine a flow over time: spot a step that produced weak output, adjust the flow, and re-run. Each new run adds another record, so you always have a trail of how the flow behaved across executions.