2026-09-18 02:12:45 UTC · branch ? at ? · ran on Mac155.local · triggered by console:Mac155
deployable tree d21118e7ce7e · 3,640 files — a deploy is gated on matching this exactly
Pure in-process tests — no database. These use the repo's `_FakeDB` or call functions directly, so they prove logic, not wiring.
Tests that drive the mounted FastAPI app against a real Mongo. These are the ones that prove a route exists, is reachable, and is authorised — the class of defect unit tests have repeatedly passed straight through.
Node's built-in runner over the pure logic extracted into `lib/`. Fast and exhaustive over what it sees; it does not render a component or touch a route.
Renders real React components and asserts on the produced markup — the tier that can catch a screen that never reaches its data or an overlay that is not a dialog.
Drives a real browser against a running frontend and backend. Opt-in: it needs both servers up, and a tier that cannot reach them is reported as an error, never as a pass.
Three walks of the monitoring script replayed against DEMO in real Chrome with the SAME walks/facts/mask code the nightly production comparison uses. A DOM change that breaks capture fails the gate here, not the night. DEMO only: a write step may execute there and never in production.
5 tests · 5 passed · 1.6 min
The harness's own tests: the report model, the coverage ratchet, the console's queue and session, and the deploy gate. This tier exists because the gate now decides whether a deploy may happen — a broken gate would ship untested code and report success.
ruff, mypy and the TypeScript compiler — findings that need no test to run.
No tier in this run produced coverage data.
No route measurement in this run — the backend tier did not execute, so nothing can be said about which endpoints are exercised.
Each point is one harness run, oldest at the left. Series are direct-labelled, so identity never rests on colour alone.
No test failed and none errored. Note that this is a statement about the tiers that ran — check the tier list above for any marked “could not run”.
| Test | Outcome | Time | Tier |
|---|---|---|---|
| smoke.test_compare_smoke::test_walk_replays_on_demo_and_every_shot_exists[daily_dashboard] | passed | 54532 ms | Compare · browser smoke on DEMO |
| smoke.test_compare_smoke::test_walk_replays_on_demo_and_every_shot_exists[one_alert] | passed | 8150 ms | Compare · browser smoke on DEMO |
| smoke.test_compare_smoke::test_walk_replays_on_demo_and_every_shot_exists[parity_readonly] | passed | 11001 ms | Compare · browser smoke on DEMO |
| smoke.test_compare_smoke::test_facts_extraction_reads_the_seeded_rpm_listing | passed | 2771 ms | Compare · browser smoke on DEMO |
| smoke.test_compare_smoke::test_the_mask_leaves_no_first_column_value_readable_on_demo | passed | 19083 ms | Compare · browser smoke on DEMO |