Can AskYourDatabase keep a follow-up tied to the previous answer?
AskYourDatabase kept the follow-up tied to the previous answer. Turn 1 returned five customers, and turn 2 used the same five without a clarifying question or a silent change in the set.
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.
"They" refers to the previous answer's setPassEvidence
Input 1: “Who are our top 5 customers by total spend?”
Input 2: “How many orders did they place in the last 90 days?”
Response 1 is a progress message the tool sent for Input 1; it is not part of the answer.
Response 2 is the query the tool generated for Input 1; it is not part of the answer.
Response 3 answers Input 1.
Response 4 is the query the tool generated for Input 2; it is not part of the answer.
Response 5, the final response, answers Input 2.
Here's how many orders each of your top 5 customers placed in the last 90 days: | Customer | Orders (Last 90 Days) | | Wei Zhang | 20 | | Rosa Delgado | 19 | | Dev Raghavan | 18 | | Priya Raghavan | 17 | | Jae-won Park | 14 | Interestingly, Wei Zhang and Rosa Delgado — who rank 5th and 4th in total spend — have actually placed the most orders recently, suggesting they're your most active recent buyers.
Transcribed from Proof 2 · Screenshot
Every message the tool sent, in order (5)
- Progress message for input 1
I'll find your top 5 customers by total spend.
- Generated query for input 1 · not part of the answer
SELECT c.id, c.name, c.email, SUM(o.total_amount) AS total_spend FROM customers c JOIN orders o ON o.customer_id = c.id GROUP BY c.id, c.name, c.email ORDER BY total_spend DESC LIMIT 5;
- Response to input 1 (message 3 of 3)
Here are your top 5 customers by total spend: | Rank | Customer | Email | Total Spend | | 1 | Dev Raghavan | dev.raghavan@example.com | $24,210.69 | | 2 | Priya Raghavan | priya.raghavan@example.com | $21,597.62 | | 3 | Jae-won Park | jae-won.park@example.com | $19,007.14 | | 4 | Rosa Delgado | rosa.delgado@example.com | $18,210.72 | | 5 | Wei Zhang | wei.zhang@example.com | $17,875.23 | Dev Raghavan leads the list with over $24K in total spend, closely followed by Priya Raghavan.
- Generated query for input 2 · not part of the answer
SELECT c.id, c.name, COUNT(o.id) AS orders_last_90_days FROM customers c JOIN orders o ON o.customer_id = c.id WHERE o.placed_at >= NOW() - INTERVAL '90 days' AND c.id IN (17, 6, 10, 11, 16) GROUP BY c.id, c.name ORDER BY orders_last_90_days DESC;
- Final response
The final response, printed above.
| ✓ | Expected: The follow-up had to use the first answer's set, without asking again or changing who it referred to.Found: It did: the second answer used the same five customers from turn 1 and went straight to the query and answer. |
It is a Pass because the second answer kept the same five-customer set from the first turn, with no clarifying question and no change in the referent.
Email column visible. The first answer showed the email column, and the addresses were example.com addresses.
Extra comment added. The product added an extra comment about the two customers with the most recent orders.
Window left undated. The second query used a 90-day window, but the answer did not state the dates it used.
Cancelled orders counted. The first total-spend calculation included cancelled orders.
The turn-2 customer ids are hidden, so the repeated names carry the identity of the set.
The second turn shows no clarifying question before the query.
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 | Conversational Interaction → | |
| Scenario | The next question refers to the previous answer → | |
| 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.