Does Anomaly AI keep a report’s meaning when it is rerun after data changes?
Anomaly AI refreshed the saved report from 24 to 27 orders, matching the new database count of 27. It preserved the original reporting period after three orders were added within it.
1 of 1 test case passed
Every test case in this scenario has a result.
The test case
Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on this test case.
Saved report re-run after new rowsPassEvidence
“Show me the number of orders placed in the last 30 days, as of September 2, 2026.”
Original chat answer 24 orders were placed in the last 30 days as of September 2, 2026 - August 4 through September 2, inclusive, using your Asia/Calcutta timezone. This includes cancelled orders and all other statuses. Data last synced October 30, 2026. Order-count answer 24 orders were placed from August 4 through September 2, 2026, inclusive, using Asia/Calcutta local dates. Includes all statuses, including cancelled orders. Data last synced October 30, 2026. Dashboard before refresh Scenario 55 - Orders Placed in the Last 30 Days as of September 2 2026 / 30-day order count as of September 2, 2026 / 24 orders were placed from August 4 through September 2, 2026, inclusive, using Asia/Calcutta local dates. Includes all statuses, including cancelled orders. Data last synced October 30, 2026. / bar labelled 24 Dashboard after refresh Last updated: Nov 2, 2026 09:15 PM / 27 orders were placed from August 4 through September 2, 2026, inclusive, using Asia/Calcutta local dates. Includes all statuses, including cancelled orders. Data last synced October 30, 2026. / bar labelled 27 Original chat answer after refresh 24 orders were placed in the last 30 days as of September 2, 2026 - August 4 through September 2, inclusive, using your Asia/Calcutta timezone. This includes cancelled orders and all other statuses. Data last synced October 30, 2026.
Transcribed from Proof 1 · Screen recording at 0:09
Transcribed from Proof 1 · Screen recording at 0:48
Transcribed from Proof 1 · Screen recording at 1:00
| ✓ | Expected: After refresh, the report should show 27 orders for the same relative period, August 4 through September 2, 2026.Found: The report refreshed from 24 to 27 after three orders were added, matching database counts of 24 then 27; August 4 through September 2, 2026 stayed unchanged. |
With the question’s as-of date held fixed, refreshing incorporated the new orders without shifting the reporting period. That meets the requirement to recalculate rather than retain a snapshot, on the plan used.
Manual refresh needed. The saved dashboard did not update automatically or warn that its total was stale. The customer had to use its database refresh action.
Narrative updates too. The dashboard’s sentence beneath the title updated along with its chart.
Chat keeps the old answer. The original order-count chat still said 24 while the refreshed dashboard said 27, leaving the conversation out of step with the saved report.
Configuration and setup
How this tool was set up for the run and what the test needed in place. Each row is a fact from the run's records; a fact the records do not hold is left out, not guessed.
How this scenario is graded
How we decide Pass, Fail and Not gradable. The same rules apply to every tool tested on this scenario.
Each test case gets one result per tool: Pass, Fail or Not gradable. A test case we haven't run yet shows Untested. There are no partial results.
- Pass: the tool did everything the test expected, and nothing it said contradicts the correct answer.
- Fail: at least one expected behaviour clearly didn't happen; the test case's result says which one and quotes the tool.
- Not gradable: our evidence couldn't settle the outcome (for example a record we needed is missing). It is never counted as a fail, and the test case's result says what's missing.
Where this sits in the benchmark
This page is one cell of a larger study: one tool, one scenario. Only this benchmark's frame appears here.
| Level | Name | Scope |
|---|---|---|
| Benchmark | AI Database Agents → | 28 scenarios · 11 tools |
| Capability | Reporting → | |
| Scenario | The report is run again after the data has changed → | |
| Tool | Anomaly AI → |
Global scenario definition → · Global capability definition → · Anomaly AI product page →
History of this result
What has happened to this result since it was first published. Runs and grades are never overwritten: a retest or a re-grade publishes a new result and keeps the earlier one readable.
Act on this result
Nothing filed here edits the run or the grade. A challenge opens a review, and a review can produce a new run or a re-grade — which becomes the current result and leaves this one in the history.
You run the same kind of test against your own setup and get the same behaviour.
Agree →Yours behaves differently. Tell us what you got, with a screenshot if you have one.
Disagree →Something here is wrong — a reference value, a transcription, a grade.
Report an issue →Tell us what changed and we schedule a rerun of the failing test case. The old result stays as history.
Vendor notice →The same record is available as structured data through the AI Demos MCP server, with the counts, the coverage and every per-test-case reason carried as fields.
Verify the proof files
These files support this result. Open a file to inspect the original evidence.
File fingerprints (SHA-256)
A fingerprint identifies the exact file used for this result.