Can Dify trace an answer claim back to supporting source text?
Dify passed this citation check. The answer said acknowledgement would happen within 90 minutes by a named duty engineer, and the top citation pointed to that same passage. The other citations also came from the same policy.
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 counts this row.
Citation opened and read against the claim it supportsPassEvidence
“How quickly does Fenwake acknowledge a Priority 1 fault reported outside business hours?”
"conversation_id": "e7869607-60d2-4863-a446-d0adec0feae2", "mode": "chat", "answer": "Within 90 minutes of the ticket being raised, by a named duty engineer.",
Raw request and response show the answer, four citations, and the run settings.
Copied from Captured file (text), lines 18–20
Open the text file ↗{ "position": 1, "dataset_id": "084fc635-481f-4970-bf09-299c83eca3c2", "dataset_name": "R6-managed-rag-ss-v1", "document_id": "0eeae560-0e74-44e9-aaee-2cc6b599dee8", "document_name": "A01_field_service_policy_and_charges_uk.pdf", "data_source_type": "upload_file", "segment_id": "7b0e6aea-05b6-4ff2-8a7b-3b6bccc09d8c", "retriever_from": "api", "score": 0.6682562, "hit_count": null, "word_count": null, "segment_position": null, "index_node_hash": null, "content": "Fenwake Instruments Ltd - Field Service Policy and Charges (United Kingdom)\r\nFenwake internal - service operations Page 4\r\n6. Out-of-hours cover\r\nOutside business hours the service desk is not staffed. A duty engineer rota covers every hour of the year,\r\nand the out-of-hours number in section 8 pages the engineer on call.\r\n6.1 Acknowledgement out of hours\r\nA Priority 1 fault reported outside business hours is acknowledged by a named duty engineer within 90\r\nminutes of the ticket being raised.", "page": null, "doc_metadata": { "region": "uk" }, "title": null, "files": null, "summary": null }, { "position": 2, "dataset_id": "084fc635-481f-4970-bf09-299c83eca3c2", "dataset_name": "R6-managed-rag-ss-v1",
Copied from Captured file (text), lines 24–50
| ✓ | Expected: The citation should point to text that supports the answer’s actual claim, not just the same document or topic.Found: The top citation points to page 4 section 6.1, which states the same 90-minute acknowledgement claim. |
The citation was opened against the answer it supported. The answer said acknowledgement would happen within 90 minutes by a named duty engineer, and the top citation pointed to that same passage.
Third citation page 2. The third citation points to the business-hours definition on page 2.
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 row says which 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 row 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 | Managed RAG → | 15 scenarios · 7 tools |
| Capability | Citations → | |
| Scenario | A claim in the answer traced back to its source → | |
| Tool | Dify → |
Global scenario definition → · Global capability definition → · Dify 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.