SeqAuto - improve browsing/QC UI (most-used, least-improved area per SA Path logs)

🤖 Written by Claude

*(Updated: the original evidence table here was mistakenly drawn from VG4 dev/test box logs. Replaced with real VG3 production figures.)*

Analysis of SA Path **VG3 production** nginx logs (4–18 Aug 2026, 15 days, ~170k human requests) shows the seqauto run pages are in daily lab use, while almost all recent seqauto work has been API/ingest plumbing for the VG4 migration — the pages people actually look at have had essentially no feature work since ~2022.

### VG3 production usage (15 days)

| Page / feature | Requests |
|---|---|
| Sequencing runs list + grid (`/seqauto/sequencing_runs`, grid handler/config) | ~700 |
| `view_sequencing_run` + `view_sequencing_run_stats_tab` (across Haem, TSO, RhampIDHplus, WGS runs) | ~120 |
| `api/view_enrichment_kit_summary` | 289 |
| EnrichmentKit / SequencingRun autocompletes | ~290 |
| SeqAuto home | ~200 |

The automated seqauto API ingest (~1,200 requests) runs continuously behind this.

### Suggestion

Focus improvements on the **run list, run detail and stats tabs** — that is where the traffic is. The software-versions pages and QC graph pages are essentially unused in VG3 production (single digits over 15 days), so they are not the priority. Ask the SA Path users what they'd most want from the run pages; the VG4 migration is the natural moment.

No specific file, test, or acceptance criterion is named. Begin by reviewing the sequencing-runs list/grid, view_sequencing_run, and view_sequencing_run_stats_tab entry points, then confirm priorities with SA Path users. The issue is ready only once a concrete run-page improvement and completion criteria are agreed.

Written by the indexing model from the issue text.