Does BlazeSQL report analytics that match what actually happened?
BlazeSQL failed to report analytics that matched what actually happened. Query review showed only one logged question, while the two chat lists showed five questions across both accounts and the walkthrough found no section with an aggregate count of questions asked.
1 of 1 test case failed
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.
Reported totals include the unansweredFailEvidence
Input 1: “As of September 2, 2026, what's our total revenue in the last 30 days?”
Input 2: “What's our most common payment method?”
Input 3: “How many complaints have we received on social media this month?”
Input 4: “How many customers currently have an active subscription?”
Input 5: “What's our net promoter score for last quarter?”
The tool's reply is shown in the evidence under Proofs.
Proof 4 · Screen recording · ▶ 0:04–0:09; ▶ 0:11–0:15
| ✗ | Expected: The reported total counts all five questions, including the unanswered ones.Found: Only one question appeared in Query review, and the page gave no answered/unanswered breakdown. |
The product had to count all five questions, including the unanswered ones. It only showed one query and no answered/unanswered breakdown, so the reported total did not match what was asked.
Selective log. One of the five questions appears in Query review, so the product logs some activity.
Sidebar contrast. The other account's sidebar shows many more chats than the counter does, so the shortfall is broader than these five questions.
Review page role. The listed row is marked not verified and low difficulty, which shows the page is for query review.
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 | Analytics & Observability → | |
| Scenario | Reported analytics match what actually happened → | |
| 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.