Kenshiki Pulse Lab

Real device rows, unmodified from the lake. Times render in your browser's local timezone. Currently: rest-pattern detection (floor = pure order statistics, extension = physics-bounded discharge slope, not a hard uncertainty cutoff). The lake has ~40 sensor types; the story-registry architecture (see CLAUDE.md) is meant to grow into a general QA lens over all of them, not stay rest-pattern-only. Nothing here ports to Swift until it looks right against the raw signal.

Local file

Load a JSON dump directly from disk — never uploaded anywhere, stays in this browser tab.

Cloud storage

D1 is the query source and loads automatically; R2 retains each unmodified upload as its audit copy. This lab has no sign-in boundary.

Dataset ID

D1 datasets

R2 archives

Upload a new pull

Appends new rows to the saved D1 dataset and archives the unmodified JSON in R2. Existing row ids are left untouched.

Kenshiki Pulse Lab

Active data

Synthesis

Timeline (local time)

Resolved decision math

Diagnostics

Local file

Cloud storage

D1 datasets

R2 archives

Upload a new pull