Lookups keep getting enriched from zhihu.com which injects junk
Author: helicalchrisCreated Oct 12, 2025Updated Oct 12, 2025
I am running research tasks, and they are using duckduckgo, but then when it gets to the web research stage it seems to disappear off and query zhihu. That in turns seems to generate a lot of Chinese character text in the activity, and I cannot see why or where this comes from, not do I want it as it seems to add junk (I was asking about countries to test, in this case Saudi Arabia, and its injected information about healing for autism and types of lies?!)
The extract from langSmith I see is pasted below:
Source: special和specific在同时表示"专门的，特定的"时，用法上有什么 …
URL: https://www.zhihu.com/question/566412162
Most relevant content from source: the difference between specific and special lies in that specific is explicit or definite while special is distinguished by a unique or unusual quality. special: She is a special girl. (she is unique, …
Full source content limited to 1000 tokens:
Source: 如何区分specific special 和specialized的意思和使用? - 知乎
URL: https://www.zhihu.com/question/623453063
Most relevant content from source: ⒋There is no specific healing for autism. 自闭症无特效药。 第二个是 形容词特殊的 和 名词特价（常用于美国、加拿大英语）、特色节目。 ⒈Xiao Hong has mastered the special recipe for …
Full source content limited to 1000 tokens:
Source: 假期快乐，specific和particular区别在哪里？ - 知乎
URL: https://www.zhihu.com/question/473295900
Most relevant content from source: The words "special" and "specific" are often used interchangeably, but there are some subtle differences in meaning between them. Both words can be used to describe people or things …I am running default settings as cloned, but with Ollama running mistral-7b-instruct (but I saw this on other models too) and with a LANGSMITH API KEY. Otherwisee direct from git clone.
Full logs from command line pasted below - it seems to be looking at bings and other search engines too?
2025-10-12T21:56:39.345486Z [info ] Using langgraph_runtime_inmem [langgraph_runtime] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=MainThread
2025-10-12T21:56:39.350268Z [info ] Using auth of type=noop [langgraph_api.auth.middleware] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=MainThread
2025-10-12T21:56:39.351938Z [info ] Starting In-Memory runtime with langgraph-api=0.4.39 and in-memory runtime=0.14.1 [langgraph_runtime_inmem.lifespan] api_variant=local_dev langgraph_api_version=0.4.39 langgraph_runtime_inmem_version=0.14.1 thread_name=asyncio_0 version=0.4.39
2025-10-12T21:56:39.365220Z [info ] Starting metadata loop [langgraph_api.metadata] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=MainThread
2025-10-12T21:56:39.365314Z [info ] Getting auth instance: None [langgraph_api.auth.custom] api_variant=local_dev langgraph_api_version=0.4.39 langgraph_auth=None thread_name=MainThread
2025-10-12T21:56:39.367479Z [info ] Starting thread TTL sweeper with interval 5 minutes [langgraph_api.thread_ttl] api_variant=local_dev interval_minutes=5 langgraph_api_version=0.4.39 strategy=delete thread_name=asyncio_1
2025-10-12T21:56:39.521175Z [info ] HTTP Request: POST https://api.smith.langchain.com/v1/metadata/submit "HTTP/1.1 403 Forbidden" [httpx] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=MainThread
2025-10-12T21:56:39.521522Z [warning ] LangChain metadata submission failed. [langgraph_api.metadata] api_variant=local_dev error="Client error '403 Forbidden' for url 'https://api.smith.langchain.com/v1/metadata/submit'\nFor more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/403" langgraph_api_version=0.4.39 thread_name=asyncio_1
2025-10-12T21:56:39.521696Z [warning ] All metadata submissions failed, will retry [langgraph_api.metadata] api_variant=local_dev attempted=['langchain'] failed=['langchain'] langgraph_api_version=0.4.39 thread_name=asyncio_1
2025-10-12T21:56:39.673534Z [info ] Registering graph with id 'ollama_deep_researcher' [langgraph_api.graph] api_variant=local_dev graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 thread_name=asyncio_1
2025-10-12T21:56:39.700618Z [info ] Starting 1 background workers [langgraph_runtime_inmem.queue] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=asyncio_0
2025-10-12T21:56:39.700933Z [info ] Worker stats [langgraph_runtime_inmem.queue] active=0 api_variant=local_dev available=1 langgraph_api_version=0.4.39 max=1 thread_name=asyncio_1
Server started in 1.26s
2025-10-12T21:56:40.099201Z [info ] Server started in 1.26s [browser_opener] api_variant=local_dev langgraph_api_version=0.4.39 message='Server started in 1.26s' thread_name='Thread-2 (_open_browser)'
Opening Studio in your browser...
2025-10-12T21:56:40.099769Z [info ] Opening Studio in your browser... [browser_opener] api_variant=local_dev langgraph_api_version=0.4.39 message=' Opening Studio in your browser...' thread_name='Thread-2 (_open_browser)'
URL: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024
2025-10-12T21:56:40.100204Z [info ] URL: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024 [browser_opener] api_variant=local_dev langgraph_api_version=0.4.39 message='URL: https://smith.langchain.com/studio/?baseUrl=http://127.0.0.1:2024' thread_name='Thread-2 (_open_browser)'
2025-10-12T21:56:40.202102Z [info ] Queue stats [langgraph_runtime_inmem.queue] api_variant=local_dev langgraph_api_version=0.4.39 max_age_secs=None med_age_secs=None n_pending=0 n_running=0 thread_name=asyncio_0
2025-10-12T21:56:59.548106Z [info ] Created run [langgraph_api.models.run] after_seconds=0 api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 checkpoint_id=None if_not_exists=reject langgraph_api_version=0.4.39 method=POST multitask_strategy=rollback path=/threads/{thread_id}/runs/stream request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_create_ms=2 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 run_put_ms=0 stream_mode=['debug', 'messages'] stream_resumable=False temporary=False thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=MainThread
2025-10-12T21:57:00.248021Z [info ] Starting background run [langgraph_api.worker] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a resumable=False run_attempt=1 run_creation_ms=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 run_queue_ms=699 run_started_at=2025-10-12T21:57:00.247562+00:00 run_stream_start_ms=0 temporary=False thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_0
2025-10-12T21:57:00.617728Z [warning ] Failed to multipart ingest runs: langsmith.utils.LangSmithError: Failed to POST https://api.smith.langchain.com/runs/multipart in LangSmith API. HTTPError('403 Client Error: Forbidden for url: https://api.smith.langchain.com/runs/multipart', '{"error":"Forbidden"}\n') [langsmith.client] api_variant=local_dev langgraph_api_version=0.4.39 thread_name='Thread-1 (tracing_control_thread_func)'
2025-10-12T21:57:00.628339Z [info ] 2 changes detected [watchfiles.main] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=MainThread
2025-10-12T21:57:06.069987Z [info ] HTTP Request: POST http://localhost:11434/api/chat "HTTP/1.1 200 OK" [httpx] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=generate_query request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
result: content='{\n "query": "AI development and implementation in Saudi Arabia 2025",\n "rationale": "This query will provide information about the current state of AI, its development, and implementation in Saudi Arabia as of October 12, 2025."\n}' additional_kwargs={} response_metadata={'model': 'mistral:7b-instruct', 'created_at': '2025-10-12T21:57:09.737731Z', 'done': True, 'done_reason': 'stop', 'total_duration': 9414261792, 'load_duration': 2721272584, 'prompt_eval_count': 193, 'prompt_eval_duration': 3014813250, 'eval_count': 63, 'eval_duration': 3676076959, 'model_name': 'mistral:7b-instruct'} id='run--6901bef6-4f5c-449a-b032-879f52d7476f' usage_metadata={'input_tokens': 193, 'output_tokens': 63, 'total_tokens': 256}
2025-10-12T21:57:09.760859Z [warning ] /Users/[email protected]/Containers/local-research/local-deep-researcher/src/ollama_deep_researcher/utils.py:189: RuntimeWarning: This package (`duckduckgo_search`) has been renamed to `ddgs`! Use `pip install ddgs` instead.
with DDGS() as ddgs:
[py.warnings] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
2025-10-12T21:57:10.055551Z [info ] response: https://www.bing.com/search?q=AI+development+and+implementation+in+Saudi+Arabia+2025 200 [primp] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
2025-10-12T21:57:10.778696Z [warning ] Failed to send compressed multipart ingest: langsmith.utils.LangSmithError: Failed to POST https://api.smith.langchain.com/runs/multipart in LangSmith API. HTTPError('403 Client Error: Forbidden for url: https://api.smith.langchain.com/runs/multipart', '{"error":"Forbidden"}\n') [langsmith.client] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=ThreadPoolExecutor-0_0
2025-10-12T21:57:12.377402Z [info ] HTTP Request: POST http://localhost:11434/api/chat "HTTP/1.1 200 OK" [httpx] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=summarize_sources request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
2025-10-12T21:57:26.342640Z [warning ] /Users/[email protected]/.cache/uv/archive-v0/9Q1EagTg9xx78xPa4WNR_/lib/python3.11/site-packages/langgraph_api/api/assistants.py:77: LangGraphDeprecatedSinceV10: `config_schema` is deprecated. Use `get_context_jsonschema` for the relevant schema instead. Deprecated in LangGraph V1.0 to be removed in V2.0.
config_schema = graph.config_schema() # type: ignore[deprecated]
[py.warnings] api_variant=local_dev langgraph_api_version=0.4.39 method=GET path=/assistants/{assistant_id}/schemas thread_name=MainThread
2025-10-12T21:57:26.352024Z [warning ] <frozen os>:679: ResourceWarning: unclosed <socket.socket fd=22, family=2, type=1, proto=6, laddr=('127.0.0.1', 56697), raddr=('127.0.0.1', 11434)>
[py.warnings] api_variant=local_dev langgraph_api_version=0.4.39 method=GET path=/assistants/{assistant_id}/schemas thread_name=MainThread
2025-10-12T21:57:31.577223Z [warning ] /Users/[email protected]/.cache/uv/archive-v0/9Q1EagTg9xx78xPa4WNR_/lib/python3.11/site-packages/langgraph_api/api/assistants.py:77: LangGraphDeprecatedSinceV10: `config_schema` is deprecated. Use `get_context_jsonschema` for the relevant schema instead. Deprecated in LangGraph V1.0 to be removed in V2.0.
config_schema = graph.config_schema() # type: ignore[deprecated]
[py.warnings] api_variant=local_dev langgraph_api_version=0.4.39 method=GET path=/assistants/{assistant_id}/schemas thread_name=MainThread
2025-10-12T21:57:32.218622Z [warning ] Failed to send compressed multipart ingest: langsmith.utils.LangSmithError: Failed to POST https://api.smith.langchain.com/runs/multipart in LangSmith API. HTTPError('403 Client Error: Forbidden for url: https://api.smith.langchain.com/runs/multipart', '{"error":"Forbidden"}\n') [langsmith.client] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=ThreadPoolExecutor-0_0
2025-10-12T21:57:35.302864Z [info ] HTTP Request: POST http://localhost:11434/api/chat "HTTP/1.1 200 OK" [httpx] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=reflect_on_summary request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_0
2025-10-12T21:57:39.840455Z [info ] Worker stats [langgraph_runtime_inmem.queue] active=1 api_variant=local_dev available=0 langgraph_api_version=0.4.39 max=1 thread_name=asyncio_1
2025-10-12T21:57:40.341513Z [info ] Queue stats [langgraph_runtime_inmem.queue] api_variant=local_dev langgraph_api_version=0.4.39 max_age_secs=datetime.datetime(2025, 10, 12, 21, 56, 59, 547988, tzinfo=datetime.timezone.utc) med_age_secs=datetime.datetime(2025, 10, 12, 21, 56, 59, 547988, tzinfo=datetime.timezone.utc) n_pending=0 n_running=1 thread_name=asyncio_1
result: content='{\n "knowledge_gap": "The summary does not provide specific details about the types of AI technologies being implemented in Saudi Arabia, their technical aspects, or any emerging trends.",\n "follow_up_query": "What are the specific AI technologies being implemented in Saudi Arabia\'s healthcare sector and financial services? Also, what are the current trends and advancements in these areas?"\n}' additional_kwargs={} response_metadata={'model': 'mistral:7b-instruct', 'created_at': '2025-10-12T21:57:40.379543Z', 'done': True, 'done_reason': 'stop', 'total_duration': 8815116083, 'load_duration': 14946250, 'prompt_eval_count': 630, 'prompt_eval_duration': 3714625041, 'eval_count': 85, 'eval_duration': 5081162584, 'model_name': 'mistral:7b-instruct'} id='run--9cfbaf1e-74f6-4b97-bf09-c84fa0dd6f55' usage_metadata={'input_tokens': 630, 'output_tokens': 85, 'total_tokens': 715}
2025-10-12T21:57:40.392833Z [warning ] /Users/[email protected]/Containers/local-research/local-deep-researcher/src/ollama_deep_researcher/utils.py:189: RuntimeWarning: This package (`duckduckgo_search`) has been renamed to `ddgs`! Use `pip install ddgs` instead.
with DDGS() as ddgs:
[py.warnings] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
2025-10-12T21:57:40.421627Z [warning ] /Users/[email protected]/.cache/uv/archive-v0/9Q1EagTg9xx78xPa4WNR_/lib/python3.11/site-packages/langgraph_api/api/assistants.py:77: LangGraphDeprecatedSinceV10: `config_schema` is deprecated. Use `get_context_jsonschema` for the relevant schema instead. Deprecated in LangGraph V1.0 to be removed in V2.0.
config_schema = graph.config_schema() # type: ignore[deprecated]
[py.warnings] api_variant=local_dev langgraph_api_version=0.4.39 method=GET path=/assistants/{assistant_id}/schemas thread_name=MainThread
2025-10-12T21:57:41.044846Z [warning ] Failed to send compressed multipart ingest: langsmith.utils.LangSmithError: Failed to POST https://api.smith.langchain.com/runs/multipart in LangSmith API. HTTPError('403 Client Error: Forbidden for url: https://api.smith.langchain.com/runs/multipart', '{"error":"Forbidden"}\n') [langsmith.client] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=ThreadPoolExecutor-0_0
2025-10-12T21:57:43.663300Z [info ] response: https://www.bing.com/search?q=What+are+the+specific+AI+technologies+being+implemented+in+Saudi+Arabia%27s+healthcare+sector+and+financial+services%3F+Also%2C+what+are+the+current+trends+and+advancements+in+these+areas%3F 200 [primp] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
2025-10-12T21:57:44.260006Z [info ] HTTP Request: GET https://www.zhihu.com/question/566412162 "HTTP/1.1 403 Forbidden" [httpx] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
Warning: Failed to fetch full page content for https://www.zhihu.com/question/566412162: Client error '403 Forbidden' for url 'https://www.zhihu.com/question/566412162'
For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/403
2025-10-12T21:57:44.633017Z [info ] HTTP Request: GET https://www.zhihu.com/question/623453063 "HTTP/1.1 403 Forbidden" [httpx] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
Warning: Failed to fetch full page content for https://www.zhihu.com/question/623453063: Client error '403 Forbidden' for url 'https://www.zhihu.com/question/623453063'
For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/403
2025-10-12T21:57:45.010718Z [info ] HTTP Request: GET https://www.zhihu.com/question/473295900 "HTTP/1.1 403 Forbidden" [httpx] api_variant=local_dev assistant_id=a6ab75b8-fb3d-5c2c-a436-2fee55e33a06 graph_id=ollama_deep_researcher langgraph_api_version=0.4.39 langgraph_node=web_research request_id=caa640b9-e897-43dd-9306-3a3a40dc9d5a run_attempt=1 run_id=0199da6d-4a1b-740c-b26c-09007f146cb5 thread_id=04f7da63-211c-4317-9a76-f443ac02c765 thread_name=asyncio_1
Warning: Failed to fetch full page content for https://www.zhihu.com/question/473295900: Client error '403 Forbidden' for url 'https://www.zhihu.com/question/473295900'
For more information check: https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/403
Warning: No raw_content found for source https://www.zhihu.com/question/566412162
Warning: No raw_content found for source https://www.zhihu.com/question/623453063
Warning: No raw_content found for source https://www.zhihu.com/question/473295900
2025-10-12T21:57:45.195751Z [warning ] /Users/[email protected]/.cache/uv/archive-v0/9Q1EagTg9xx78xPa4WNR_/lib/python3.11/site-packages/langgraph_api/api/assistants.py:77: LangGraphDeprecatedSinceV10: `config_schema` is deprecated. Use `get_context_jsonschema` for the relevant schema instead. Deprecated in LangGraph V1.0 to be removed in V2.0.
config_schema = graph.config_schema() # type: ignore[deprecated]
[py.warnings] api_variant=local_dev langgraph_api_version=0.4.39 method=GET path=/assistants/{assistant_id}/schemas thread_name=MainThread
2025-10-12T21:57:45.694604Z [warning ] Failed to send compressed multipart ingest: langsmith.utils.LangSmithError: Failed to POST https://api.smith.langchain.com/runs/multipart in LangSmith API. HTTPError('403 Client Error: Forbidden for url: https://api.smith.langchain.com/runs/multipart', '{"error":"Forbidden"}\n') [langsmith.client] api_variant=local_dev langgraph_api_version=0.4.39 thread_name=ThreadPoolExecutor-0_0
2Source: langchain-ai/local-deep-researcher