Can Extend AI add one member and remove another in a repeated set?
Extend AI let a user add and remove invoice items through Human Review. After approval, the traffic cones were present, the site fencing was absent and the telehandler and generator values were unchanged.
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.
Add one member and remove onePassEvidence
In Human Review, we removed the site fencing and added traffic cones, then saved the corrections and approved the result.
{ "line_items": [ { "amount": 850, "quantity": 1, "unit_price": 850, "description": "Telehandler, 4-tonne, weekly hire" }, { "amount": 420, "quantity": 2, "unit_price": 210, "description": "Generator, 20kVA, weekly hire" }, { "amount": 45, "quantity": 3, "unit_price": 15, "description": "Traffic cones, set of 20, weekly hire" } ] }
Reviewed JSON: the saved set with traffic cones and without site fencing
Copied from Output file (JSON), lines 1–22
Open the JSON file ↗| ✓ | Expected: The added member must be saved, the removed member absent, and every other member unchanged.Found: The saved set contained the required 3 traffic-cone sets (£45), no site fencing, and unchanged entries for 1 telehandler (£850) and 2 generators (£420). |
- line_items 1
- { "amount": 850, "quantity": 1, "unit_price": 850, "description": "Telehandler, 4-tonne, weekly hire" }
- line_items 2
- { "amount": 420, "quantity": 2, "unit_price": 210, "description": "Generator, 20kVA, weekly hire" }
- line_items 3
- { "amount": 192, "quantity": 8, "unit_price": 24, "description": "Site fencing, per 10m panel, weekly" }
Show the whole file (1 KB)
{ "line_items": [ { "amount": 850, "quantity": 1, "unit_price": 850, "description": "Telehandler, 4-tonne, weekly hire" }, { "amount": 420, "quantity": 2, "unit_price": 210, "description": "Generator, 20kVA, weekly hire" }, { "amount": 192, "quantity": 8, "unit_price": 24, "description": "Site fencing, per 10m panel, weekly" } ] }
- line_items 1
- { "amount": 850, "quantity": 1, "unit_price": 850, "description": "Telehandler, 4-tonne, weekly hire" }
- line_items 2
- { "amount": 420, "quantity": 2, "unit_price": 210, "description": "Generator, 20kVA, weekly hire" }
- line_items 3
- { "amount": 45, "quantity": 3, "unit_price": 15, "description": "Traffic cones, set of 20, weekly hire" }
- line_items
- [ { "amount": 850, "quantity": 1, "unit_price": 850, "description": "Telehandler, 4-tonne, weekly hire" }, { "amount": 420, "quantity": 2, "unit_price": 210, "description": "Generator, 20kVA, weekly hire" }, { "amount": 45, "quantity": 3, "unit_price": 15, "description": "Traffic cones, set of 20, weekly hire" } ]
Human Review, rather than the playground, is the product's correction screen. It saved both membership changes without altering the other items, satisfying both required operations; the result shown is the reviewed set within Extend AI, not the downstream webhook payload.
Playground editing limits. The Extract playground's Form view offered no way to add or delete rows, including when a row was under the cursor. These edits were available in Human Review.
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 | Structured Document Extraction → | 19 scenarios · 10 tools |
| Capability | Review and Correction → | |
| Scenario | Repair a record → | |
| Tool | Extend AI → |
Global scenario definition → · Global capability definition → · Extend 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.