Can FutureSmart Database Agent keep using a corrected metric definition?
FutureSmart Database Agent kept the corrected revenue definition on the repeated question and the reworded one. After the customer gave the rule in chat, both later answers excluded cancelled orders and returned USD 51,848.02 from one row.
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: “Our revenue definition excludes cancelled orders — please use that going forward.”
Input 2: “As of September 2, 2026, what was our total revenue over the last 90 days?”
Input 3: “As of September 2, 2026, how much did we earn in total over the last 90 days?”
Response 1 replies to Input 1.
Response 2 replies to Input 2.
Response 3 replies to Input 3.
Response to input 1 Got it—I will exclude cancelled orders from the revenue definition going forward. Response to input 2 Total Revenue (Last 90 Days, Excluding Cancelled Orders) Period: 90 days ending on September 2, 2026 Currency: USD Total Revenue: USD 51,848.02 This figure represents the sum of all order amounts placed within the specified 90-day window, with any orders marked as cancelled excluded from the calculation. Response to input 3 Total Earned Over the Last 90 Days (as of September 2 2026) USD 51,848.02 in revenue was generated from orders that were not cancelled during the 90-day period ending on September 2, 2026. This figure represents the summed total_amount of all qualifying orders, grouped by currency.
Transcribed from Proof 1 · Screenshot
Transcribed from Proof 2 · Screenshot
Transcribed from Proof 3 · Screenshot
| ✓ | Expected: Both follow-up questions use the corrected definition and match the reference answer.Found: Both later answers excluded cancelled orders and returned USD 51,848.02, matching the total computed directly from the database for the same 90-day window. |
The customer first gave the rule in chat. On both later questions, the answers excluded cancelled orders, returned one row, and showed USD 51,848.02, so the corrected definition held for both prompts.
Different SQL, same result. The two follow-up questions used slightly different SQL but the same filter and returned the same row.
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 | FutureSmart Database Agent → |
Global scenario definition → · Global capability definition → · FutureSmart Database Agent 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.