status

This status skill renders an experiment dashboard for an autoresearch workflow, summarizing results, active loops, and progress across experiments. It answers the question "how is my experiment going?" by aggregating run counts, kept results, best metrics, percentage change, and status into a readable table, invoked through the `/ar:status` command.

The skill delegates its work to a bundled Python script (`log_results.py`) that can report on a single experiment, a whole domain, or a full dashboard across all experiments. For each experiment it also inspects a `loop.json` file to surface any active recurring loop, showing its interval, cron ID, and start date. Results can be exported to CSV or Markdown by passing a format flag and an output path. The example output is a plain aligned table with columns for domain, experiment, runs, kept, best value, change, status, and loop cadence.

It is aimed at researchers and engineers running iterative optimization or A/B-style experiments who want a quick read-only snapshot of experiment health and progress, or an exportable report to share. The operations it performs are limited to reading experiment state files and printing or exporting a summary.