Does RAGFlow report when an uploaded file cannot be processed?
RAGFlow did not flag the two unreadable uploads as document failures. A truncated PDF and a password-protected PDF received the same success status as a readable comparison document.
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.
Unprocessable file submitted through the normal upload pathFailEvidence
We uploaded a readable comparison document, a truncated PDF and a password-protected PDF, then started parsing through RAGFlow’s API.
Only the tool's reply to inputs 1, 2 and 3 is copied here; its reply to input 4 was not copied as text. The rest of what the tool returned is shown in the evidence under Proofs.
{'run': 'DONE', 'status': '1', 'progress': 1.0, 'chunk_count': 0, 'token_count': 0, 'size': 2114, 'process_duration': 9.9354}
API response log: comparison document, repeated ingestion statuses, stored copies and retrieval results
Copied from Captured file (text), line 43 · an excerpt
Open the whole text file ↗Everything the tool returned, in order (5)
- Truncated PDF status
{'run': 'DONE', 'status': '1', 'progress': 1.0, 'chunk_count': 0, 'token_count': 0, 'size': 2114, 'process_duration': 7.49227}
- Password-protected PDF status
The final response, printed above.
- Truncated PDF ingestion log
11:51:22 Page(1~2): No chunk built from bad_01_truncated.pdf
- Password-protected PDF ingestion log
11:51:24 Page(1~2): No chunk built from bad_02_password_protected.pdf
- Good control document status
{'run': 'DONE', 'status': '1', 'progress': 1.0, 'chunk_count': 2, 'token_count': 567, 'size': 2775, 'process_duration': 3.72094}
Source excerpt · partial file · response 3
11:51:22 Page(1~2): No chunk built from bad_01_truncated.pdf 11:51:22 Page(1~2): Task done (6.01s)
Copied from Captured file (text), lines 36–37 · an excerpt
Source excerpt · partial file · response 4
11:51:24 Page(1~2): No chunk built from bad_02_password_protected.pdf 11:51:24 Page(1~2): Task done (9.01s) uploaded bytes: 174963 | stored bytes: 2114 stored copy head: b'%PDF-1.7\n%\xc7\xec\x8f\xa2\n%%Invocation: gs ? ? -sDEVICE=pdfwrite -dPDFS' | length 2114 | contains Ghostscript: True scoped retrieval probe: code 0 chunks 0
Copied from Captured file (text), lines 52–56 · an excerpt
| ✗ | Expected: Report a failure for each document that cannot be processed.Found: Both PDFs were marked successfully completed with zero chunks, including in the existing dataset. Failure appeared in free-text progress messages, not as a document failure status. |
Both unreadable PDFs received success rather than failure statuses. A readable comparison document produced 2 chunks through the same upload and parsing path, so this is a reporting failure, not an execution failure. This conclusion covers REST API status; the web console’s document list was not inspected.
Stored rewrite. Both unreadable uploads were stored as 2,114-byte Ghostscript rewrites, and neither the status nor the response said the original file had been replaced.
Zero-chunk retrieval. A retrieval probe for each unreadable file returned success with zero chunks, so neither file contributed retrievable content.
Failure state exists. The product can return a failure state on this path, but these uploads did not receive it.
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 | Managed RAG → | 15 scenarios · 9 tools |
| Capability | Data Ingestion → | |
| Scenario | A bad file in the upload → | |
| Tool | RAGFlow → |
Global scenario definition → · Global capability definition →
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.