Inspect runs and reports
List runs and open their ranked reports and receipts.
1. List
Get a list of your runs to find the one you want to inspect.
Open Proofs in the main nav. All your runs appear as cards with the experiment name, date, winner, and a link to the ranked report.
Not available via the CLI. Use the app's Proofs page to list runs.
MCP has no tool that lists past runs. list_experiments lists your experiment containers (id, name, objective) but not their runs:
{}Keep the run_id that prove_task or run_experiment returned, or open Proofs in the app to browse runs by experiment.
2. Open a report
The ranked report shows each candidate's quality score, cost, latency, and savings. Expand any candidate to read the per-item receipts.
Click a run in Proofs to open the ranked report. Click any candidate row to expand the receipts drill-down, which shows the model's output and score for each input item.
Write the full receipts to a local file when you run an eval:
redcrown eval task.json --report-json out.jsonOpen out.json to inspect per-item outputs and scores.
Fetch a run summary with get_run or the full ranked report with get_report:
{
"run_id": "..."
}On a run with repeat_n of 2 or more, each item receipt also carries three per-item stability fields:
quality_by_pass, the item's own score on each pass, in pass order.spread, the widest gap between any two of those scores.other_passes, the item's output and score on every pass after the first, present only on an item whose spread crossed the 0.05 unstable threshold. This lets a reader see what changed instead of taking the spread on faith. Only the widest-changing items per candidate carry this field; every item still keeps its ownquality_by_passandspread.
A missing or absent value on any of the three means stability was not measured for that item. The usual causes are a run that did not repeat, or an item whose result was reused across passes rather than re-measured. It is never a signal that the item was perfectly stable.
Last verified .