Can AskYourDatabase answer questions from connected data?
AskYourDatabase answered three connected-data questions correctly, but missed the 30-day order count. It returned 24 instead of 25 for that lookup, while it got the month difference, August total, and top product by units right.
3 of 4 test cases passed · 1 failed
Every test case in this scenario has a result.
The 4 test cases
Each test case is judged on its own: Pass, Fail, Not gradable, or Untested. The scenario result above is based on these test cases.
Order count in a defined periodFailEvidence
“As of September 2, 2026, how many orders were placed in the last 30 days?”
As of September 2, 2026, 24 orders were placed in the last 30 days (from August 3 through September 1, 2026).
Transcribed from Proof 1 · Screenshot
Every message the tool sent, in order (3)
- Progress
I'll query the database to find orders placed in the last 30 days before September 2, 2026.
- Generated SQL · not part of the answer
SELECT COUNT(*) AS order_count FROM orders WHERE placed_at >= '2026-08-03' AND placed_at < '2026-09-02'
- Answer
The final response, printed above.
| ✗ | Expected: 25 orders in the 30 days ending 2 September 2026.Found: It gave 24 orders; the reference count for the connected data was 25. |
| ✗ | Expected: The count should cover 3 August through 2 September 2026, including the as-of date.Found: It said the window ran from 3 August through 1 September, so 2 September was left out. |
It counted through 1 September instead of through 2 September, so the 2 September order was left out and the answer came back one low.
Window shown plainly. The answer states the date window in plain words, so the counted period is visible without opening the SQL.
Count label visible. The SQL block makes the count label visible.
Order-count difference between two monthsPassEvidence
“As of September 2, 2026, how many more orders were placed in August 2026 than in July 2026?”
Answer
In August 2026, there were 3 more orders placed than in July 2026.
Order counts in the answer
Month | Orders Placed; July 2026 | 23; August 2026 | 26
Answer restatement
August had 26 orders compared to July's 23 — a difference of 3 orders.
Transcribed from Proof 1 · Screenshot
Every message the tool sent, in order (5)
- Progress
I'll compare order counts between July and August 2026.
- Generated SQL · not part of the answer
SELECT COUNT(*) FILTER (WHERE placed_at >= '2026-07-01' AND placed_at < '2026-08-01') AS july_orders, COUNT(*) FILTER (WHERE placed_at >= '2026-08-01' AND placed_at < '2026-09-01') AS august_orders FROM orders;
- Answer
The final response, printed above.
- Order counts in the answer
The final response, printed above.
- Answer restatement
The final response, printed above.
| ✓ | Expected: August 2026 should have 3 more orders than July 2026: 26 versus 23.Found: It gave 3 more orders, matching the reference difference of 3 (26 in August and 23 in July). |
| ✓ | Expected: August 2026 should have more orders than July 2026: 26 versus 23.Found: It said August had more orders than July, and the table shows 26 versus 23. |
It compared the two full months and gave August the higher count, with a difference of 3 orders.
Three result forms. The answer shows the difference, both month counts, and the direction in a sentence, a table, and a closing line.
Total order value in a calendar monthPassEvidence
“As of September 2, 2026, what was the combined total amount of all orders placed in August 2026?”
The combined total amount of all orders placed in August 2026 was $13,441.94.
Transcribed from Proof 1 · Screenshot
Every message the tool sent, in order (3)
- Progress
I'll query the database to find the total amount of orders placed in August 2026.
- Generated SQL · not part of the answer
SELECT COALESCE(SUM(total_amount), 0) AS combined_total FROM orders WHERE placed_at >= '2026-08-01' AND placed_at < '2026-09-01'
- Answer
The final response, printed above.
| ✓ | Expected: $13,441.94 combined across all August 2026 orders.Found: It gave $13,441.94, matching the reference total of $13,441.94 for the connected data. |
| ✓ | Expected: It should cover August 2026 and include all orders.Found: Its SQL covers all August orders and the answer says all orders placed in August 2026. |
It summed all August orders with no status filter and returned the expected total of $13,441.94.
Period in one line. The answer gives the total and repeats the month, but the status breakdown is only visible in the SQL.
Top product by units ordered in a monthPassEvidence
“As of September 2, 2026, which product had the most units ordered in August 2026, and how many units?”
As of September 2, 2026, the product with the most units ordered in August 2026 was the Soy candle trio, with 5 units ordered.
Transcribed from Proof 1 · Screenshot
Every message the tool sent, in order (3)
- Progress
I'll find which product had the most units ordered in August 2026.
- Generated SQL · not part of the answer
SELECT p.name, SUM(oi.quantity) AS total_units FROM order_items oi JOIN orders o ON o.id = oi.order_id JOIN products p ON p.id = oi.product_id WHERE o.placed_at >= '2026-08-01' AND o.placed_at < '2026-09-01' GROUP BY p.name ORDER BY total_units DESC LIMIT 1;
- Answer
The final response, printed above.
| ✓ | Expected: A top product for August 2026: Brass picture frame 8x10, Outdoor rug 6x9, or Soy candle trio, tied at 5 units.Found: It named Soy candle trio, one of three products tied at the reference maximum of 5 units. |
| ✓ | Expected: 5 units, the highest product total for August 2026.Found: It gave 5 units, the maximum on the data it queried. |
It summed August units across products and returned Soy candle trio with 5 units, which matches one of the tied maxima.
Single top product. The answer names one product only, so the tie is not visible from the message alone.
No tie breaker. The SQL stops after one row without a tie-breaker, so a tied winner can vary.
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 | Question Answering → | |
| Scenario | The answer is available in the connected data → | |
| Tool | AskYourDatabase → |
Global scenario definition → · Global capability definition → · AskYourDatabase 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.