### What problem does this PR solve?
Fix [Bug]: Save parser configs in dataset configuration page is not
working #15175
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
## Summary
Fixes the confirmed asyncio anti-patterns from #14755. Only the three
verified bugs are addressed; patterns already correctly using
`asyncio.new_event_loop()` in a fresh thread are left untouched.
### Changes
**`api/apps/restful_apis/tenant_api.py` — fire-and-forget
`send_invite_email`**
`asyncio.create_task()` was called without storing the `Task` reference.
CPython's GC can collect an unfinished task, silently cancelling it and
swallowing exceptions. Fixed by storing the task in a module-level
`_background_tasks: set[Task]` with a `done_callback` to discard it on
completion — the standard Python idiom for safe background tasks.
**`api/apps/restful_apis/agent_api.py` — fire-and-forget
`background_run`**
Same root cause in the webhook "Immediately" execution path. Same fix
applied.
**`rag/llm/chat_model.py` (`LocalLLM._stream_response`) —
`asyncio.get_event_loop()` on running loop**
`asyncio.get_event_loop()` returns Quart's running event loop when
called from an async context.
Calling `loop.run_until_complete()` on it raises `RuntimeError`.
Replaced with `asyncio.new_event_loop()` so the generator
uses a dedicated fresh loop, closed in a `finally` block.
## What was NOT changed
- `llm_service._sync_from_async_stream` and
`evaluation_service._sync_from_async_gen`: both already correctly use
`asyncio.new_event_loop()` inside a fresh thread.
- `llm_service._run_coroutine_sync`: only caller is `rag/app/resume.py`
(sync context), so `thread.join()` is correct there.
- `requests` in agent tools: sync methods dispatched through thread
pools; httpx migration is a separate, larger refactor.
## Test plan
- [ ] Invite a team member and confirm the email is sent with no task
warnings in logs.
- [ ] Trigger a webhook agent in "Immediately" mode; confirm canvas
state is persisted after background run.
- [ ] Verify `LocalLLM` (Jina backend) chat and streaming work
end-to-end.
Closes#14755
---------
Co-authored-by: Zhichang Yu <yuzhichang@gmail.com>
Follow on PR #15146 to reslove the backwad compatability issue.
1. /agents/<attachment_id>/download ->
/agents/attachments/<attachment_id>/download
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
## Summary
This change fixes ingestion quality issues where MinerU parser output
may contain HTML fragments (for example, table-related tags like `<tr>`,
`<td>`, `<br>`), which were previously passed directly into
chunking/tokenization and degraded chunk quality.
The fix adds a sanitization step in the MinerU parser path so parsed
sections are normalized to clean text before chunking.
## Change Type (select all)
- [x] Bug fix
- [x] Ingestion pipeline improvement
- [x] Parser/chunking quality fix
## Related Issue
- https://github.com/infiniflow/ragflow/issues/14831
### What problem does this PR solve?
This PR improves the table upload flow for CSV/Excel files by allowing
table column role configuration at upload time.
Previously, users had to:
1. Upload and parse a table file.
2. Open parser settings and manually set table column roles.
3. Re-parse the file for the roles to take effect.
This was inefficient and required an unnecessary second parse.
With this change:
1. When the knowledge base uses table parsing, the upload dialog
extracts CSV/Excel headers client-side.
2. Users can choose Auto mode or Manual mode.
3. In Manual mode, users can assign per-column roles before upload.
4. The selected parser config is sent with the upload request and
applied server-side during document creation.
Result: configured table column roles are applied from the first parse.
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
Co-authored-by: Ahmad Intisar <ahmadintisar@Ahmads-MacBook-M4-Pro.local>
### What problem does this PR solve?
1. Fix /chat/completions to send only the latest message
2. Allo chat stream=False
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
Fix /chat/completions not aware of conversation_id
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
Fix: /openai/<chat_id>/chat/completions not aware of session_id
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
move agent attachment download api to the correct route and update
frontend callers
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### Notes
- Move the attachment download endpoint from document routes to agent
routes.
- Update frontend download callers to use the agent attachment endpoint.
- Reuse the shared file response header helper instead of duplicating it
in `agent_api.py`.
## Summary
Implements the TODO in `evaluation_service.py`: **Track token usage** in
evaluation results.
## Changes
- **Import** `num_tokens_from_string` from `common.token_utils`
- **Prompt tokens**: Use the full prompt returned by `async_chat` when
available (includes system prompt + knowledge base + query), otherwise
fall back to the question token count
- **Completion tokens**: Count tokens in the generated answer
- **Storage**: Store `token_usage` as `{prompt_tokens,
completion_tokens, total_tokens}` in each `EvaluationResult` instead of
`None`
## Why
The evaluation pipeline previously saved `token_usage: None` for every
result. This change allows downstream consumers (e.g. evaluation
dashboards, cost tracking) to see approximate token usage per test case
using the same tokenizer (tiktoken cl100k_base) used elsewhere in
RAGFlow.
## Testing
- No new tests added; existing evaluation flow unchanged
- Token counting uses existing `num_tokens_from_string` utility
---------
Co-authored-by: kiannidev <kiannidev@users.noreply.github.com>
### What problem does this PR solve?
The agent API currently does not pass chat_template_kwargs to the
underlying LLM call path, so clients cannot control template-level model
behavior (such as thinking-mode toggles) when invoking
/agents/chat/completion. This PR adds passthrough support for
chat_template_kwargs across agent execution flows (session and
non-session, streaming and non-streaming) by propagating it through
canvas runtime state and into LLM invocation kwargs. This addresses the
feature gap raised in [Issue
#14182](https://github.com/infiniflow/ragflow/issues/14182).
Closes#14182
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
Closes#14789
### What problem does this PR solve?
User API endpoints (`login`, `user_profile`, `user_add`,
`forget_reset_password`) were returning full user objects via
`to_json()` / `to_dict()`, which included sensitive fields like
`password` and `access_token` in the response body. This leaks
credentials to the client.
This PR adds a `to_safe_dict()` method on the `User` model that strips
sensitive fields (`password`, `access_token`) and replaces all affected
call sites to use it.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
1. Enhance retry and timeout, and adjust the default timeout
2. NER: spacy do not batch chunks
3. extract _has_cancel_and_exit
4. enhance log messages
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
- [x] Refactoring
`GET /agents/<agent_id>/sessions/<session_id>` crashed with
`AttributeError: 'NoneType' object has no attribute 'to_dict'` when the
session lookup failed: `_, conv =
API4ConversationService.get_by_id(...)` returned `(False, None)`, then
`conv.to_dict()` was called unconditionally.
This is reachable in multi-instance deployments: the session row may not
yet be visible on the node servicing the immediate follow-up GET after a
session is created on a different node.
Add the same `if not exists` guard already used by every other call site
of `API4ConversationService.get_by_id` (see agent_api.py:1147,
sdk/session.py:179, conversation_service.py:248, canvas_service.py:323).
Closes#14989
### What problem does this PR solve?
_Briefly describe what this PR aims to solve. Include background context
that will help reviewers understand the purpose of the PR._
### Type of change
- [ ] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):
### What problem does this PR solve?
Closes#15048.
Several SDK session routes in `api/apps/sdk/session.py` called
`.split()` directly on `request.headers.get("Authorization")`. When
clients omitted the header, the handlers raised `AttributeError` before
returning the existing `Authorization is not valid!` response.
This PR centralizes SDK Authorization parsing in a small helper and
keeps the existing error response for missing, empty, or malformed
headers.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### Tests
- `ZHIPU_AI_API_KEY=dummy uv run --python 3.13 --group test pytest
test/testcases/test_http_api/test_session_management/test_session_sdk_routes_unit.py::test_sdk_session_routes_missing_authorization_unit
-q`
- `uv run --python 3.13 --group test ruff check api/apps/sdk/session.py
test/testcases/test_http_api/test_session_management/test_session_sdk_routes_unit.py`
- `python3 -m py_compile api/apps/sdk/session.py
test/testcases/test_http_api/test_session_management/test_session_sdk_routes_unit.py`
- `git diff --check`
### What problem does this PR solve?
Remove duplicate function definitions in
`api/db/services/dialog_service.py`.
**Problem:** Two helper functions were defined twice in the same file,
but with different parameter orders:
- First definition (line 57):
`_resolve_reference_metadata(request_payload=None, config=None)`
- Second definition (line 136): `_resolve_reference_metadata(config,
request_payload=None)`
**Solution:** Keep the second definition (which is actually used by
other modules) and remove the first one to avoid confusion.
Additionally, remove duplicate `_enrich_chunks_with_document_metadata`
definition (keep line 140 version).
<img width="1584" height="313" alt="image"
src="https://github.com/user-attachments/assets/7daee832-244f-4bb2-8488-e3b65012a3f9"
/>
<img width="1672" height="359" alt="image"
src="https://github.com/user-attachments/assets/4fd2f523-273c-4b20-a7c9-ab35740b7834"
/>
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [x] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):
## Summary
- Align **GET `/api/v1/documents/<doc_id>/download`** with
**`/preview`**: resolve extension and MIME type from the stored document
name when the **`ext` query parameter is omitted**, instead of
defaulting to `markdown`.
- When **`?ext=`** is present, behavior stays the same as before
(explicit extension / `Content-Type` mapping).
- Enforce the same access + document lookup pattern as preview
(**`accessible`** + **`get_by_id`**).
- Extend unit tests for the no-`ext` PDF filename case.
## Test plan
- [x] `uv run pytest
test/testcases/test_web_api/test_document_app/test_document_metadata.py::TestDocumentMetadataUnit::test_download_attachment_success_and_exception_unit`
- [x] Optional: `curl -sSI` against
`/api/v1/documents/<pdf_doc_id>/download` without `ext` and confirm
`Content-Type: application/pdf`
Fixes#15052.
POST /api/v1/dify/retrieval resolved the caller via @apikey_required
(injecting tenant_id) but then fetched the requested knowledge_id with
no tenant filter and ran the full retrieval pipeline against
kb.tenant_id (the owner). Any valid Dify-compatible API key could
retrieve chunks from any tenant whose KB UUID was known. Adds the
missing ownership check.
## Root Cause
api/apps/sdk/dify_retrieval.py line 253:
KnowledgebaseService.get_by_id(kb_id) fetched the KB by id alone, then
the handler used kb.tenant_id (the OWNER) to build the embedding model
and call the retriever. The caller tenant_id was only used downstream at
line 278 for retrieval_by_children, well after cross-tenant data was
already retrieved.
grep confirmed there was no KnowledgebaseService.accessible call
anywhere in the handler.
## Fix
Two-line guard immediately after the existing get_by_id lookup,
mirroring the pattern PR #14749 lands for the sibling sdk/doc.py routes
(download, parse, stop_parsing, retrieval_test):
e, kb = KnowledgebaseService.get_by_id(kb_id)
if not e:
return build_error_result(message="Knowledgebase not found!",
code=RetCode.NOT_FOUND)
+ if not KnowledgebaseService.accessible(kb_id, tenant_id):
+ return build_error_result(message="No authorization.",
code=RetCode.AUTHENTICATION_ERROR)
if kb.tenant_embd_id:
...
KnowledgebaseService.accessible already handles solo-tenant ownership,
team membership via TenantService.get_joined_tenants_by_user_id, and the
permission=ME distinction. No behavior change for legitimate callers;
cross-tenant callers now receive RetCode.AUTHENTICATION_ERROR (109).
## Test Plan
- [x] Regression test added:
test/unit_test/api/apps/sdk/test_dify_retrieval.py
- test_cross_tenant_request_is_rejected -- attacker tenant calling owner
tenant KB gets 109; retriever is not invoked
- test_same_tenant_request_succeeds -- owner tenant gets the records
back
- test_missing_knowledge_base_returns_not_found -- missing KB returns
404 BEFORE the access check fires (legit callers see the clearer
message)
- [x] All 3 tests pass after the fix
- [x] Cross-tenant test FAILS on pre-fix main (KeyError on result[code]
because handler leaks records dict instead of returning auth error)
- [x] ruff check clean on both changed files
- [x] No drive-by reformatting in dify_retrieval.py -- only the 2 added
lines
### Post-fix output
test_cross_tenant_request_is_rejected PASSED [ 33%]
test_same_tenant_request_succeeds PASSED [ 66%]
test_missing_knowledge_base_returns_not_found PASSED [100%]
============================== 3 passed in 0.04s
===============================
Closes#15027
### What problem does this PR solve?
Closes#15076
Two endpoints in `api/apps/restful_apis/chat_api.py` accepted a
`user_id` field from the request body and used it directly when creating
a session:
```python
# before (vulnerable)
"user_id": req.get("user_id", current_user.id) # create_session
conv = await _create_session_for_completion(chat_id, dia, req.get("user_id", current_user.id)) # session_completion
```
Any authenticated caller could supply an arbitrary `user_id` and have
the new session attributed to a different user — effectively spoofing
session ownership. Both call sites are now fixed to always use
`current_user.id`, which is set by the authentication middleware and
cannot be tampered with via the request payload.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### Changes
| File | Change |
|------|--------|
| `api/apps/restful_apis/chat_api.py` | Remove `req.get("user_id", ...)`
fallback in `create_session` and `session_completion`; always use
`current_user.id` |
|
`test/testcases/test_http_api/test_session_management/test_session_sdk_routes_unit.py`
| Add `test_create_session_user_id_not_spoofable` and
`test_session_completion_user_id_not_spoofable` (both `@pytest.mark.p2`)
|
### Testing
Two new unit tests assert that a `user_id` value supplied in the request
body is silently ignored and the session is always owned by the
authenticated user:
```
test_create_session_user_id_not_spoofable
test_session_completion_user_id_not_spoofable
```
Run with:
```bash
uv run pytest test/testcases/test_http_api/test_session_management/test_session_sdk_routes_unit.py -k "spoofable" -v
```
### What problem does this PR solve?
Closes#15025
Langfuse-enabled `dialog_service.async_chat()` regressed to
`langfuse_tracer.start_generation(...)` after the earlier Langfuse v4
migration. Langfuse v4 uses `start_observation(as_type="generation")`,
so the remaining `start_generation` call can fail when chat tracing is
enabled.
This restores the migrated `start_observation(as_type="generation")`
call for chat observations while preserving the existing trace context,
model, input payload, and update/end flow. It also adds a regression
test with a fake Langfuse v4-style client that exposes
`start_observation()` but not `start_generation()`.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### Tests
- `.venv/bin/pytest
test/unit_test/api/db/services/test_dialog_service_final_answer.py -q`
- `.venv/bin/ruff check api/db/services/dialog_service.py
test/unit_test/api/db/services/test_dialog_service_final_answer.py`
## Summary
Closes#14921.
Reconfiguring an existing LLM provider to enable **tool call** or
**vision** fails with `Your API key is invalid. Fail to access model.`
even when the saved API key is correct. The most visible report is
VLLM ("Cannot add vllm model" once `--enable-auto-tool-choice` /
vision is toggled on), but the bug applies to every provider whose
api_key field stays blank in edit mode.
## Root cause
PR #14885 ("Fix: llm add api key overridden") removed the existing-key
lookup in `api/apps/llm_app.py::add_llm`. The intent was correct —
stop the saved key from clobbering a user-provided new one — but the
removal was unconditional, so the edit path now has no fallback at all:
1. `web/src/pages/user-setting/setting-model/hooks.tsx:230` sets the
initial `api_key` form value to `''` in edit mode (the real key is
never returned to the browser).
2. The user toggles `is_tools` / `vision` without retyping the key.
3. `hooks.tsx:183-185` strips the empty `api_key` from the payload.
4. `add_llm` defaults to the placeholder `"x"`
(`api/apps/llm_app.py:182`).
5. The upstream provider rejects `"x"` with `Your API key is invalid`.
## Fix
Restore the fallback **narrowly**, before any factory-specific handler
runs:
- If `req.get("api_key") is None`, look up the tenant's existing record
(using the correctly suffixed `llm_name` for VLLM /
OpenAI-API-Compatible / LocalAI / HuggingFace).
- Decode the saved blob with `_decode_api_key_config` and write **only
the decoded `api_key` string** back into `req["api_key"]`. Never use
the raw JSON payload — that was the exact thing PR #14885 was trying
to avoid.
- When the user **does** type a new key, `req.get("api_key")` is not
`None` and the fallback is skipped, so PR #14885's fix is preserved.
| Scenario | Before this PR | After this PR |
|---|---|---|
| Plain factory (VLLM, Ollama, …), retype key | OK | OK |
| Plain factory, blank key in edit (the bug) | Fails with "API key is
invalid" | Recovers saved key, validates against the real one |
| OpenRouter / Bedrock, change `provider_order` only | Fails |
`apikey_json([...])` rebuilds the JSON with saved `api_key` + new field
|
| User clears the form and types a brand-new key | OK (key replaced) |
OK (key replaced — fallback skipped) |
## Files changed
- `api/apps/llm_app.py` — restored fallback in `add_llm` (no other call
sites touched).
## Test plan
- [ ] Add a VLLM chat model with a valid api_key, no toggles → save
succeeds.
- [ ] Edit the same model, toggle **tool call** on, leave api_key blank
→ save succeeds, validation runs against the saved key.
- [ ] Edit again, toggle **vision** on (model_type → `image2text`),
leave api_key blank → save succeeds.
- [ ] Edit again and **type a new api_key** → the new key replaces the
saved one (`is None` check skips the fallback). Verify via the DB
row or by deliberately typing a wrong key and observing the
validation failure.
- [ ] Repeat the blank-key edit with **OpenRouter**, changing only
`provider_order` → resulting api_key JSON contains the saved
`api_key` and the new `provider_order`.
- [ ] First-time add of a new model name → no existing record, fallback
no-ops, behaves as before.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [ ] New Feature (non-breaking change which adds functionality)
- [ ] Documentation Update
- [ ] Refactoring
- [ ] Performance Improvement
- [ ] Other (please describe):
## What problem does this PR solve?
Closes#12017.
TTS output is deterministic for a given `(model, text)` pair, so
re-running the same text through the same TTS model produces the same
bytes — yet `Canvas.tts` and `dialog_service.tts` re-synthesized on
every request. That's slow and wastes provider quota whenever the same
assistant response is replayed, shared across users, or repeated within
a session.
### Change
New helper `rag/utils/tts_cache.py` with `synthesize_with_cache(tts_mdl,
cleaned_text)`:
- **Key:** `tts:cache:{model_id}:{sha256(text)}` — separate namespace
per model, identical cleaned text reuses a single entry across both call
sites.
- **Value:** the hex-encoded audio blob both call sites already
returned. No format change for downstream consumers.
- **TTL:** 7 days by default, configurable via
`RAGFLOW_TTS_CACHE_TTL_SECONDS`.
- **Failure modes:** a Redis hiccup falls back to direct synthesis; a
failed synthesis still returns `None` (existing contract preserved).
[`Canvas.tts`](https://github.com/infiniflow/ragflow/blob/main/agent/canvas.py#L683-L724)
and
[`dialog_service.tts`](https://github.com/infiniflow/ragflow/blob/main/api/db/services/dialog_service.py#L1367-L1380)
now route through the helper; the per-file bytes-accumulation/hex-encode
loop has been removed in favor of one shared implementation.
## Type of change
- [x] New Feature (non-breaking change which adds functionality)
## Test plan
- [ ] **Cache hit, chat path:** Configure a dialog with TTS enabled, ask
the same question twice with `stream=false`. Verify the second response
returns the same `audio_binary` and that the second invocation doesn't
hit the TTS provider (e.g., observe provider-side logs / usage counters;
check no `LLMBundle.tts can't update token usage` log line on the second
run).
- [ ] **Cache hit, agent path:** Same exercise via a Conversational
Agent that includes a Message component playing back the answer.
- [ ] **Cache isolation per model:** Switch tenant's `tts_id` between
two models, run the same text against each — confirm the second model's
first synthesis still happens (no cross-model hits).
- [ ] **TTL override:** Set `RAGFLOW_TTS_CACHE_TTL_SECONDS=120`, confirm
the entry expires after 2 minutes.
- [ ] **Redis unavailable:** Stop Redis (or break the connection).
Verify the TTS endpoint still works — synthesis falls back to direct
calls, with a `TTS cache lookup failed` / `TTS cache store failed`
warning logged.
- [ ] **Failure path:** Configure a TTS model with an invalid API key,
ensure the response still returns successfully with `audio_binary=None`
(no regression vs. current behavior).
### What problem does this PR solve?
Fix minor code quality issues:
1. Fix typo in assertion error message: "Can't fine" → "Can't find"
2. Remove duplicate line in common/connection_utils.py
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
- [x] Refactoring
### Related issues
Closes#14922
### What problem does this PR solve?
`POST /memories` already resolves `tenant_llm_id` and `tenant_embd_id`
through `ensure_tenant_model_id_for_params`, but `PUT
/memories/<memory_id>` accepted client-supplied `tenant_llm_id` /
`tenant_embd_id` without checking that those `tenant_llm` rows belong to
the memory owner’s tenant. A caller could persist another tenant’s row
IDs and later trigger extraction or embedding that loaded foreign model
credentials via `get_model_config_by_id(tenant_model_id)` with no tenant
allow-list.
This change aligns the update path with create: updates that change
models must go through `llm_id` / `embd_id` and
`ensure_tenant_model_id_for_params` scoped to the **memory’s**
`tenant_id` (not only the current user, so team-access cases stay
correct). Direct `tenant_*` fields in the body without `llm_id` /
`embd_id` are rejected. As defense in depth, `memory_message_service`
passes `allowed_tenant_ids` / `requester_tenant_id` into
`get_model_config_by_id` for LLM and embedding resolution so mismatched
IDs cannot be used even if bad data existed. A regression test rejects
payloads that set only `tenant_llm_id` / `tenant_embd_id`.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
---------
Co-authored-by: jony376 <jony376@gmail.com>
### What problem does this PR solve?
This PR improves the connector dashboard task management experience and
adds better visibility into connector execution logs.
### Overview:
#### Before
<img width="700" alt="image"
src="https://github.com/user-attachments/assets/e4a8ed6f-2e18-4f0f-8528-41a514550052"
/>
#### Now:
<img width="700" alt="Screenshot from 2026-05-18 16-31-30"
src="https://github.com/user-attachments/assets/d4ca193b-847a-49ae-9e4f-5fbca60ea627"
/>
### 1. Add a new logging page to the connector dashboard
A new logging page has been added so users can view connector task
execution logs directly from the connector dashboard.
### 2. Merge the Resume button into Confirm
The separate **Resume** button has been removed. The **Confirm** button
now represents different actions depending on the current task state:
- **Save**: Save form changes and reschedule tasks.
- **Stop**: Cancel currently scheduled or running tasks.
- **Resume**: Create new scheduled tasks after the previous tasks have
been stopped.
- **Start**: Start tasks when no task has been started yet.
### 3. Separate syncing and pruning tasks
Connector tasks are now separated into **syncing** and **pruning**.
Pruning is controlled by the **Sync deleted files** option:
- When **Sync deleted files** is disabled, only syncing tasks are shown.
- When **Sync deleted files** is enabled, both syncing and pruning tasks
are shown.
**Now: Sync deleted files disabled**
<img width="700" alt="Sync deleted files disabled"
src="https://github.com/user-attachments/assets/dbd9232e-614a-407f-a0b1-c109e5fa567d"
/>
**Now: Sync deleted files enabled**
<img width="700" alt="Sync deleted files enabled"
src="https://github.com/user-attachments/assets/1f527f48-ccb3-4ee8-97ca-086891489296"
/>
### 4. Update logs in backend
<img width="700" alt="image"
src="https://github.com/user-attachments/assets/10a95a3f-98c1-4e67-8afa-ddf6cda5b0b2"
/>
### 5. Remove connector resume API
- Removed: `POST /v1/connectors/<connector_id>/resume`
- Replaced by: `PATCH /v1/connectors/<connector_id>`
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
### What problem does this PR solve?
When _parse_doc_id_filter_with_metadata returns [], the empty list is
falsy so the WHERE id IN (...) clause was silently skipped, causing the
full dataset to be returned instead of an empty result.
Change `if doc_ids:` to `if doc_ids is not None:` in both get_list() and
get_by_kb_id() to distinguish between no filter (None) and a filter that
matched zero documents ([]).
Fixes#14962
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
### What problem does this PR solve?
1. expose batch_chunk_token_size for configuration
2. retrieve chunks when build subgraph for the doc, not retreive all
docs chunks at the begining
3. get all chunks for a document, used to be hard coded 10000
4. delete not used method run_graphrag
### Type of change
- [x] New Feature (non-breaking change which adds functionality)
- [x] Refactoring
Follow on: #14617
### What problem does this PR solve?
Fixes#14746.
Adds tenant access checks for connector-by-id REST routes before reading
connector details, mutating connector config/status, deleting
connectors, rebuilding, or listing sync logs. Unauthorized callers now
receive `RetCode.AUTHENTICATION_ERROR` with `No authorization.` without
reaching the connector/log mutation paths.
Validation:
- `python3 -m pytest
--confcutdir=test/testcases/test_web_api/test_connector_app
test/testcases/test_web_api/test_connector_app/test_connector_routes_unit.py`
- `uvx ruff check api/apps/restful_apis/connector_api.py
api/db/services/connector_service.py
test/testcases/test_web_api/test_connector_app/test_connector_routes_unit.py`
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
Co-authored-by: dev111-actor <dev111-actor@users.noreply.github.com>
### What problem does this PR solve?
Refactor: speed up ragflow server, save startup memory, saved 200MiB,
and 5-9 seconds start time.
##### Before
1241292 | | \_ python3 api/ragflow_server.py
RAGFlow server is ready after 25.61845850944519s initialization.
##### After
1019968 | | \_ python3 api/ragflow_server.py
RAGFlow server is ready after 16.205134391784668s initialization.
### Type of change
- [x] Refactoring
### What problem does this PR solve?
Closes#14858
The `test_db_connection` endpoint in the agent API accepts a
user-supplied `host` and connects to it directly via database drivers
(MySQL/PostgreSQL) without any validation. This allows an attacker to
probe internal network addresses (e.g. `127.0.0.1`, `10.x.x.x`,
link-local, etc.) through the server — a classic Server-Side Request
Forgery (SSRF) vulnerability.
This PR adds an SSRF guard that resolves the host and rejects any
address that is not globally routable before the database connection is
attempted.
**Changes:**
- **`common/ssrf_guard.py`** — Added `assert_host_is_safe()`, a
host-level counterpart of the existing `assert_url_is_safe()`, designed
for non-HTTP protocols (database drivers) where there is no URL to
parse.
- **`api/apps/restful_apis/agent_api.py`** — Call
`assert_host_is_safe(req["host"])` at the top of `test_db_connection` so
that non-public hosts are rejected early with a clear error message.
Fixes#14858
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
---------
Co-authored-by: Jin Hai <haijin.chn@gmail.com>
## Summary
- Stop pulling chunk vectors (`q_*_vec`) back from Elasticsearch in the
main retrieval path. ES already knows them; shipping them was pure
bandwidth/memory overhead.
- Recover the per-chunk cosine similarity via a second KNN-only ES call
filtered by the candidate chunk ids. The new `_score` is merged with
locally computed term similarity using the user-configured
`vector_similarity_weight`.
- Lazily fetch the chunk embedding only for the chunks
`insert_citations` actually needs.
## Details
**`rag/nlp/search.py`**
- `Dealer.search`: no longer appends `q_*_vec` to the ES select list.
OceanBase still gets it (its rerank path is unchanged).
- New `Dealer._knn_scores(sres, idx_names, kb_ids)`: a `MatchDenseExpr`
over the cached query vector filtered by `id IN sres.ids`, returning
`{chunk_id: cosine_score}` via ES `_score`.
- New `Dealer.rerank_with_knn(...)`: term similarity from
`qryr.token_similarity` plus the ES-supplied KNN score, combined with
`tkweight`/`vtweight` and the existing rank-feature bonus.
- New `Dealer.fetch_chunk_vectors(chunk_ids, tenant_ids, kb_ids, dim)`:
on-demand vector fetch for citation use.
- `Dealer.retrieval` routes Infinity → unchanged, OceanBase → existing
local `rerank`, ES → new KNN-score path.
**`common/doc_store/es_conn_base.py`**
- New `get_scores(res)` helper returning `{_id: _score}` directly from
hit headers (ES doesn't surface `_score` through `get_fields`).
**`api/db/services/dialog_service.py`**
- New top-level `_hydrate_chunk_vectors(...)` helper. On ES it
back-fills `ck["vector"]` from `fetch_chunk_vectors` right before
`insert_citations`. No-op on Infinity / OB (their chunks already carry
vectors).
- Both `decorate_answer` closures became `async` and are `await`-ed at
all call sites in `async_chat` and `async_ask`.
## Backend behavior
| Backend | Returns chunk vec in main search | Sim source | Vectors for
citations |
|---|---|---|---|
| ES | No | second KNN call (`_score`) merged with term sim | fetched on
demand |
| Infinity | No (unchanged) | normalized `_score` | already on chunks |
| OceanBase | Yes (kept) | local hybrid rerank | already on chunks |
## Test plan
### What problem does this PR solve?
This fixes a bug where files uploaded in chat were left in storage after
the session was deleted. It now removes those chat-uploaded blobs during
session deletion. fixes#14965
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
Closes#14753
## What changed
| File | Change |
|---|---|
| `pyproject.toml` | `requires-python` → `>=3.13,<3.15`; remove
`strenum==0.4.15` |
| `Dockerfile` | `uv python install 3.13`, `uv sync --python 3.13` |
| `.github/workflows/tests.yml` | `uv sync --python 3.13` on both matrix
legs |
| `CLAUDE.md` | dev setup command + requirements note updated |
| `deepdoc/parser/mineru_parser.py` | `from strenum import StrEnum` →
`from enum import StrEnum` |
| `agent/tools/code_exec.py` | same |
`StrEnum` has been in the stdlib since Python 3.11 — the `strenum`
backport package is no longer needed once the floor is 3.13.
## Why uv.lock is not regenerated
`uv lock --python 3.13` fails because:
1. The infiniflow/graspologic fork pins `numpy>=1.26.4,<2.0.0`
2. `tensorflow-cpu>=2.20.0` (the first release with cp313 wheels)
depends on `ml-dtypes>=0.5.1`, which requires `numpy>=2.1.0`
3. These two constraints are irreconcilable on Python 3.13
The lockfile regeneration requires loosening the `numpy` upper bound in
the `infiniflow/graspologic` fork. Once that fork commit is updated and
the SHA in `pyproject.toml:49` is bumped, `uv lock --python 3.13` will
succeed.
## RFC corrections
Two claims in the original RFC (#14753) did not hold up under code
review:
- **"graspologic hard-blocks 3.13"** — the infiniflow fork at the pinned
commit has no `<3.13` Python constraint. The blocker is the transitive
`numpy<2.0.0` conflict with tensorflow-cpu's test dependency, not a
direct Python version cap.
- **"free-threading throughput gains for I/O-bound workload"** — Python
3.13 free-threading requires a special `--disable-gil` build and
provides no benefit for async I/O code (the GIL is already released
during I/O). The real motivation is forward compatibility and improved
error messages.
## What problem does this PR solve?
Closes#13384.
The `/api/v1/agentbots/<agent_id>/completions` non-streaming path
returned the first yielded SSE chunk and exited:
```python
async for answer in agent_completion(objs[0].tenant_id, agent_id, **req):
return get_result(data=answer)
```
That meant structured output, the full assistant message, and reference
data were all dropped when an agent was called with `stream=false`.
Streaming worked because each event was forwarded individually;
non-streaming was returning a raw SSE-formatted string from a single
early event.
The v1 endpoint at
[`agent_api.py:1006-1050`](https://github.com/infiniflow/ragflow/blob/main/api/apps/restful_apis/agent_api.py#L1006-L1050)
already handles this correctly. This PR mirrors that aggregation in the
SDK beta endpoint: parse each SSE line, accumulate `content` from
`message` events, merge `reference`, collect `outputs.structured` from
each `node_finished` event keyed by `component_id`, and attach all of
them to the final response.
## Type of change
- [x] Bug fix (non-breaking change which fixes an issue)
## Test plan
- [ ] Build an agent with a node that emits structured output, call
`POST /api/v1/agentbots/<agent_id>/completions` with `stream=false` and
a beta API token, verify `data.structured.<component_id>` is present in
the response.
- [ ] Same agent with `stream=true` — verify behavior is unchanged.
- [ ] Agent without structured output — verify `data.structured` is
omitted, `content` and `reference` still aggregated correctly.
## Summary
Harden `api/utils/configs.deserialize_b64` so that it always routes
pickle data through the existing `RestrictedUnpickler`
(`restricted_loads`) rather than falling back to bare `pickle.loads()`.
- **CWE-502** — Deserialization of Untrusted Data
- **File / function**: `api/utils/configs.py` → `deserialize_b64`
- **Caller**: `SerializedField.python_value` in `api/db/db_models.py`
(invoked by Peewee whenever a pickled DB column is read)
## The issue
Before this change, `deserialize_b64` consulted a
`use_deserialize_safe_module` config flag that **defaults to `False`**
and is not set anywhere in the repository:
```python
use_deserialize_safe_module = get_base_config('use_deserialize_safe_module', False)
if use_deserialize_safe_module:
return restricted_loads(src)
return pickle.loads(src) # <-- default path
```
So the default code path was unrestricted `pickle.loads()` on bytes read
from a MySQL `SerializedField(serialized_type=PICKLE)` column. Any
attacker who can influence those bytes (SQL injection elsewhere,
compromised DB credentials, a backup restored from an untrusted source,
or a compromised replication peer) can craft a pickle payload that
achieves arbitrary code execution on the ragflow application server when
the field is next read.
Today no model in-tree instantiates a `SerializedField` with the default
PICKLE type — only `JsonSerializedField` is used in practice — so the
attack surface is currently **latent** rather than actively reachable
through an HTTP endpoint. But the insecure-by-default behaviour is a
sharp edge: any future field that uses the default PICKLE serialization
would silently inherit RCE-on-read semantics.
## The fix
```diff
- use_deserialize_safe_module = get_base_config(
- 'use_deserialize_safe_module', False)
- if use_deserialize_safe_module:
- return restricted_loads(src)
- return pickle.loads(src)
+ return restricted_loads(src)
```
`restricted_loads` is the existing `RestrictedUnpickler` already defined
in the same file, which limits permitted modules to `numpy` and
`rag_flow`. The config flag (and the now-dead `get_base_config` import)
are removed.
Diff is 1 insertion / 6 deletions, scoped to a single function.
## Testing
- Built a malicious pickle whose `__reduce__` resolves to
`posix.system('id')`. Pre-fix: executes. Post-fix: `restricted_loads`
raises `UnpicklingError: global 'posix.system' is forbidden`.
- Round-tripped a benign `numpy.ndarray` through `serialize_b64` →
`deserialize_b64`. Values preserved bit-for-bit.
- Confirmed `use_deserialize_safe_module` is not set in any config file
in the tree, so removing the flag does not change any operator-facing
knob that was actually in use.
## A note on `restricted_loads` itself
The existing `SECURITY.md` notes that `restricted_loads`'s `numpy`
allow-list can still be reached via `numpy.f2py.diagnose.run_command`.
This PR does **not** attempt to fix that — it is a separate hardening
question about tightening the allow-list to specific symbols rather than
whole modules. The change here strictly improves on the status quo (bare
`pickle.loads`) and brings the default path in line with what the
`restricted_loads` helper was clearly designed for. Happy to follow up
with a separate PR narrowing the allow-list if that direction is
welcome.
## Adversarial review
Before submitting, we tried to argue this finding away. The two
strongest objections are (1) "no field uses PICKLE today, so this is
unreachable" — true, but the default behaviour of a security-sensitive
helper still matters because new fields silently inherit it; and (2)
"the attacker already needs DB write access, which is game over" —
partially true, but pickle-RCE meaningfully escalates *data tampering*
into *code execution on the application host* (filesystem, internal
network, in-process secrets), which is not equivalent. The fix is one
line of real code, has no behavioural cost for legitimate callers, and
removes an insecure default. We decided it was worth filing.
---
<sub>_Submitted by Sebastion — autonomous open-source security research
from [Foundation Machines](https://foundationmachines.ai). Free for
public repos via the [Sebastion AI GitHub
App](https://github.com/marketplace/sebastion-ai)._</sub>
### What problem does this PR solve?
add document download endpoint and refactor existing download function
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
## Summary
- Fixes resource-management bugs in the `POST
/agents/test_db_connection` endpoint where database connections could be
left open on error (part of #14750)
## Changes
- `api/apps/restful_apis/agent_api.py` — `test_db_connection`:
- mysql / mariadb / oceanbase / postgres: replaced bare `db.connect()` /
`db.close()` fallthrough with `with db.connection_context()` and a probe
`SELECT 1` — guaranteed close on both success and exception
- mssql: nested `try/finally` blocks so `cursor.close()` and
`db.close()` are always called even when `cursor.execute()` raises
- trino: wrapped cursor ops in `try/finally` for the same reason
- Removed the `if req["db_type"] != "mssql": db.connect(); db.close()`
shared fallthrough block — each branch now owns its teardown
- Consolidated to a single `return get_json_result(...)` after the
if/elif chain
### What problem does this PR solve?
Fixes#14866.
Previously, `DocumentService.increment_chunk_num` and
`decrement_chunk_num` updated the `Document` row and its parent
`Knowledgebase` row in two separate, non-transactional statements. If
the second update failed (DB error, connection drop, etc.) after the
first one succeeded, the document and knowledge base chunk/token
counters would drift apart and stay inconsistent.
There was also a behavioral asymmetry between the two methods:
- `increment_chunk_num` only logged a warning when the document row was
missing and returned a value that callers usually treated as success.
- `decrement_chunk_num` raised `LookupError` in the same situation.
This PR makes the counter updates atomic and aligns the missing-document
behavior between the two methods:
- Wrap the `Document` and `Knowledgebase` updates in
`increment_chunk_num` / `decrement_chunk_num` inside a `DB.atomic()`
block so both succeed or both roll back together.
- Raise `LookupError` from `increment_chunk_num` when the target
document no longer exists, matching `decrement_chunk_num`.
- Update `reset_document_for_reparse` in `document_api_service.py` to
catch the new `LookupError` and return a proper "Document not found!"
API error instead of propagating the exception.
No schema changes, no API contract changes for the success path; only
the failure mode for a missing document during reparse is now a clean
error response instead of an uncaught exception.
### Type of change
- [x] Bug Fix (non-breaking change which fixes an issue)
## Summary
- Replace `gen_conf={}` with `gen_conf=None` + guard in
`rag/llm/chat_model.py` (12 instances across Base, BaiChuanChat,
LocalLLM, MistralChat, ReplicateChat, BaiduYiyanChat, GoogleChat
classes)
- Replace `doc_ids=[]` with `doc_ids=None` + guard in
`api/db/services/document_service.py` (1 instance)
- Mutable default arguments are shared across all calls, causing
potential cross-request state contamination
- See Python docs:
https://docs.python.org/3/faq/programming.html#why-are-default-values-shared-between-objects
## Test plan
- [x] Verify LLM calls work with and without explicit gen_conf
- [x] No behavior change for existing callers — `None` is replaced with
`{}` at function entry
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
Co-authored-by: Jin Hai <haijin.chn@gmail.com>
Co-authored-by: Yingfeng <yingfeng.zhang@gmail.com>
Co-authored-by: Kevin Hu <kevinhu.sh@gmail.com>
