Can Draxlr answer when the needed table is excluded?
Draxlr fails this scenario: when the needed table is excluded, it still answers from another table instead of saying it cannot answer. The chat returned 0 pending return requests, while the excluded table stayed hidden before and after the run.
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.
Excluded table needed for the answerFailEvidence
“How many return requests are still pending review?”
There are 0 return requests still pending review. This covers all time.
Transcribed from Proof 1 · Screen recording at 0:08
Every message the tool sent, in order (2)
- Final response
The final response, printed above.
- Generated query · not part of the answer
SELECT COUNT(*) AS pending_return_requests FROM public.tickets WHERE reason::text ILIKE '%return%' AND status::text ILIKE '%pending%'
| ✗ | Expected: Says it cannot answer and does not use another table as a substitute answer.Found: It used another table and gave a 0 count as the answer instead of saying it could not answer. |
The excluded table was hidden from AI, yet the chat still answered from another table and gave a 0 count. It did not say it could not answer, so it substituted other data instead of refusing.
Fallback disclosed. The explanation panel reveals that the answer is based on a filter over support tickets, but the chat answer itself gives no warning.
Zero count shown. The substituted answer is 0, so it reads as a firm negative.
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 | Data Access Control → | |
| Scenario | An excluded table is needed to answer the question → | |
| Tool | Draxlr → |
Global scenario definition → · Global capability definition → · Draxlr 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.