Can Anomaly AI use user-written documentation to reach cryptic data?
Anomaly AI used user-written documentation to answer a question about cryptically named data. It counted 7 returns flagged for manual fraud review, matching the reference count of 7.
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
“How many returns are currently flagged for manual fraud review?”
7 returns are flagged for manual fraud review, across all return statuses. This reflects the latest sync on October 30, 2026 at 11:13 AM India time.
Transcribed from Proof 1 · Screen recording at 0:08
Every message the tool sent, in order (2)
- Generated query · not part of the answer
SELECT COUNT(*) AS "value" FROM "returns" r WHERE r."flg_x2" = TRUE
- Answer
The final response, printed above.
| ✓ | Expected: After documentation, answer the question using the documented table and column.Found: It queried the documented table and review flag and returned 7 flagged returns; the reference count was also 7. |
The question named neither the table nor the column, yet Anomaly AI queried the documented data. With a new chat and conversation learning off, earlier conversations could not supply the mapping.
Meaning of currently. For the flagged-return count, Anomaly AI included returns across all statuses, not just those still awaiting action.
Return details follow-up. Below the answer, Anomaly AI offered “Show flagged return details” as a follow-up.
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 | Anomaly AI → |
Global scenario definition → · Global capability definition → · Anomaly AI 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.