Does Dot use user-written documentation for cryptic table and column names?
After a note describing the cryptically named flag column was saved, Dot answered that 7 returns are flagged for manual fraud review, counting exactly the documented column on the returns table; the reference count is 7. Without the note it had reached the same count but asked whether it had picked the right flag.
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.
Documented cryptic table and columnPassEvidence
A note titled 'Returns flg_x2 column' was saved in Dot's Model > Notes: 'flg_x2 on the returns table marks whether a return was flagged for manual fraud review — true means it needs manual review before refunding.' The question was first asked without the note (19:35 IST), then again in a new chat after it (19:41 IST); the note was deleted afterwards.
“How many returns are currently flagged for manual fraud review?”
7 returns are currently flagged for manual fraud review (flg_x2 = true).
Transcribed from Proof 1 · Screen recording at 1:40
Every message the tool sent, in order (2)
- Answer after the note was saved
The final response, printed above.
- Query Dot ran for that answer · not part of the answer
SELECT COUNT(*) FILTER (WHERE flg_x2 = TRUE) AS flagged_returns_count, COUNT(*) AS total_returns_count, ROUND(COUNT(*) FILTER (WHERE flg_x2 = TRUE) * 100.0 / NULLIF(COUNT(*), 0), 1) AS flagged_returns_PERCENT FROM customer_support_db.public.returns
| ✓ | Expected: Once the note documenting the column is saved, Dot answers the question from the documented table and column.Found: Dot answered '7 returns are currently flagged for manual fraud review', and its query counted rows where flg_x2 is true on the returns table; the reference count is 7. |
Dot had to answer from the documented table and column after the note was saved. It counted 7 flagged returns from exactly that column. Without the note it had also reached 7 but asked if that was the right flag, so the note removed its doubt, not a wrong answer.
Flag counted regardless of status. For 'currently flagged', Dot counted every flagged return whatever its status; counting only unresolved returns would give 4. Its log states no date or status exclusions.
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 database's names are cryptic and the user documents them → | |
| Tool | Dot → |
Global scenario definition → · Global capability definition → · Dot 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.