Can BlazeSQL use a user's corrected definition when answering future questions?
BlazeSQL kept the corrected revenue rule across both follow-up questions, and both answers matched the reference total of $23,469.02 from 42 orders. The saved note was already on the orders table, and the later queries excluded cancelled orders and ended the window before September 3.
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.
Corrected metric definition survives a rephrasingPassEvidence
Input 1 · Revenue definition correction: “Our revenue definition excludes cancelled orders — please use that going forward.”
Before each of Inputs 2–3: The revenue definition was held as a Suggested note on the orders table, and a new chat was opened.
Input 2 · Verbatim repeat: “As of September 2, 2026, what was our total revenue over the last 90 days?”
Input 3 · Reworded question: “As of September 2, 2026, how much did we earn in total over the last 90 days?”
Response 1 answers Input 1.
Response 2 answers Input 2.
Response 3 answers Input 3.
Acknowledgement of the correction Got it! I'll make sure that any revenue calculations going forward exclude cancelled orders (i.e., filtering out orders where status = 'cancelled'). Thanks for clarifying! Let me know if you have any queries you'd like to run. Answer to the repeat Done—revenue excludes orders cancelled as of September 2, 2026, using historical order statuses. Answer to the rewording You earned $23,469.02 USD from June 5–September 2, 2026—the 90-day period ending September 2—excluding cancelled orders.
Transcribed from Proof 1 · Screenshot
Transcribed from Proof 4 · Screenshot
Transcribed from Proof 5 · Screenshot
| ✓ | Expected: Both the repeat and the rephrasing use the corrected definition and match the reference answer.Found: Both follow-up questions used the corrected rule and each returned $23,469.02 from 42 orders. |
The case is a pass because both follow-up questions used the corrected revenue rule and matched the reference total. The saved note was already on the orders table, and both generated queries excluded cancelled orders while keeping September 2 in scope.
Saved note persisted. A saved note on the orders table carried the corrected revenue rule into this run.
Status handling differed. The repeat used historical status at the cutoff, while the reworded question used the current status column.
Window ended before cutoff. Both generated queries ended the window before September 3.
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 | Training → | |
| Scenario | The agent used the wrong definition and the user corrects it → | |
| Tool | BlazeSQL → |
Global scenario definition → · Global capability definition → · BlazeSQL 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.