Results — list view #

URL: /results

Every Litmus run lands in this table — finished and in-flight side by side — sorted by start time, newest first. The view has two parts: a stats strip above the table and the table itself.

Use it to find a specific run by serial, scan today's pass rate, drill into a failing run, or watch a station's activity live.

Stats row #

Results — stats row

When the table has rows, a card strip above the table summarises the current page's outcomes:

Stat	Meaning
Total Runs	Total run count across all runs Litmus has recorded — not just the visible page.
Pass Rate (page)	Percentage of rows on the visible page with outcome `Passed`.
Passed	Rows on the visible page with outcome `Passed` (green chip).
Failed	Rows on the visible page with outcome `Failed` (red chip).
Errored	Rows on the visible page with outcome `Errored` (amber chip).
Latest	Start time of the first row on the visible page under the current sort. Under the default sort (Started, newest first) this is the newest start; under any other sort it's the start time of whichever row appears first.

The rates and counts (except Total Runs) follow pagination — flip to the next page and the numbers reflect that page's outcomes.

Table #

Results — main table

Click a row to open the run's detail page at /results/{run_id}.

Column	What it shows	Sortable
Outcome	Colored chip — Passed, Failed, Errored, Skipped, Aborted, Terminated, Done, or Running for in-flight runs	yes
Serial	DUT serial number stamped on the run	yes
Part Number	DUT part number	yes
Hostname	Station hostname that ran the test	yes
Project	Project name from `litmus.yaml`	yes
Phase	Test phase facet (e.g. `development`, `production`, `characterization`)	yes
Started	Run start timestamp, rendered in browser-local time	yes
Steps	Total step count for the run	yes
Meas	Total measurement count for the run	yes
Ended	Run end timestamp, blank for in-flight runs	no

The table body scrolls, the header stays pinned. Click a sortable column header to sort by that column; the sort applies to all runs in the index, not just the visible page (the next page reflects the sorted order). Cells with no value (e.g. a run with no part number stamped) render blank.

The pagination footer at the bottom shows the current range (1-50 of N) and a rows-per-page selector with options of 10, 25, 50, 100, and "all."

Live updates #

The table watches the event log for run starts and ends and refreshes itself — no manual reload needed. Updates are scoped to the project's data directory (data_dir in litmus.yaml, or the default Litmus home if unset). Runs from other projects pointing at a different data directory don't appear here.

Empty state #

When the table has no rows, the stats strip is hidden and a single card appears with a "Launch a Test" button that jumps to the Launch Test view (/launch). Fresh installs always start in this state.

Underlying data #

The table reads from the runs index Litmus maintains in the background under the project's data directory. Each row corresponds to one Litmus run — the same record you get from:

litmus runs on the command line
litmus runs --json for machine-readable output
RunsQuery in the Python query API

For the full schema of one run row, see Models reference → RunSummary. For the underlying event log Litmus materialises run rows from, see Concepts → Event log.

Common tasks #

Find a flaky test — sort by Outcome, then scan for tests that appear with both Passed and Errored outcomes across recent runs.
Compare two runs from the same DUT — sort by Serial, open the two run detail pages in adjacent tabs.
Watch live activity — leave the view open during a test run; the table auto-refreshes on run.started / run.ended events.

Bookmarkable URL state #

This view does not currently encode filters or pagination in the URL — the page always opens with the default sort (Started, newest first) and the default 50-rows-per-page. Bookmarking the URL bookmarks the landing state, not the current view.