Status
How each tool is verified
This page is written when the site is built, not when you open it. Last verified means the date of this build. There is no uptime figure and no success rate here, because we do not measure either.
One row per tool
| Tool | Runs | Last verified | How it is verified | Failure state |
|---|---|---|---|---|
| AI Shelf Check | Server fetch | 84 of 84 tests passed | Unit tests in url-guard.test.ts, robots.test.ts, scoring.test.ts, controls.test.ts, config.test.ts run before every build and must all pass for the build to exist. /api/scan?url= returns the same report as JSON. The print stylesheet is the download. | A scan that cannot complete shows the reason in the form and produces no score. A page cut off at the size limit is reported as truncated, never softened. |
| Fact Split | Server fetch | 27 of 27 tests passed | Unit tests in factsplit.test.ts run before every build and must all pass for the build to exist. /api/fact-split?url= returns the same comparison as JSON. | A page that cannot be fetched shows the fetch error. No partial comparison is shown. |
| Span | Server fetch | 19 of 19 tests passed | Unit tests in span.test.ts run before every build and must all pass for the build to exist. /api/span?url=&s= returns the same stamps as JSON. | A page that cannot be fetched shows the fetch error. A paraphrase is a miss, with the nearest sentence shown and labeled not a match. |
| Falsify | In your browser | 20 of 20 tests passed | Unit tests in falsify.test.ts run before every build and must all pass for the build to exist. Export receipt (JSON) on the page, written on your device. It records inputs, outputs, tallies, timings and settings. | Without WebGPU the page says so before anything downloads. A model that fails to load shows the loader's error and no result. |
| Context Cliff | In your browser | 34 of 34 tests passed | Unit tests in contextcliff.test.ts run before every build and must all pass for the build to exist. Export receipt (JSON) on the page, written on your device. | Without WebGPU the page says so before anything downloads. A question that never held at full text is excluded from the chart rather than counted as broken. |
| Compression Receipt | In your browser | 26 of 26 tests passed | Unit tests in compressionreceipt.test.ts run before every build and must all pass for the build to exist. Export receipt (JSON) on the page, written on your device. It records a SHA-256 hash of the input, not the input. | Empty or oversized input shows a message and no table. Spans that are not in the input are listed as not scored and excluded from the total. |
| Tripwire | In your browser | 31 of 31 tests passed | Unit tests in tripwire.test.ts run before every build and must all pass for the build to exist. Export receipt (JSON) on the page, written on your device. | Without WebGPU the page says so before anything downloads. Two unreadable model replies in a row stop a run, and the raw replies are shown. |
What verified means here, and what it does not
- It means the tests passed. Before every build, the unit tests behind each tool run with vitest. If any fails, the build fails and the previous deploy stays up. The counts above are read from that run.
- It does not mean the tool was exercised live. The tests cover the pure logic: address guards, robots.txt matching, scoring, normalization, clause splitting, the agent loop with scripted replies. They do not fetch a real store, and they do not run the in-browser model.
- It is not uptime. We do not run a monitor against this site and do not publish availability. If a page is down, this page is down with it.
- It is not a success rate. No count of scans, checks or runs is kept, so no rate can be published. The receipts a visitor exports stay on the visitor's device.
- The date is the build's. Built October 7, 2026 (UTC). A newer build replaces this page and its date.
Publication protocol v1.0
- Every figure shows its value, its denominator, its window and its source.
- A figure is marked Recorded with the date of the run it came from. Nothing on this site is live telemetry.
- A tool with no run behind it says Not yet measured and names the denominator it will publish.
- Verified means the unit tests behind a tool passed when this build was made. It is not uptime and not a success rate.
- A change to how something is measured is listed on the changelog with the published figures it invalidates.
- Every measurement page states what it does not show, in its own section.