Loop detection middleware incorrectly flags legitimate scientific workflows as infinite loops, blocking continuation of complex analyses after detecting repeated tool calls. This impacts RNA-seq analysis requiring multiple sequential operations.
Loop detection should distinguish between actual infinite loops and legitimate scientific workflows requiring repeated operations. System should allow continued execution of batch processing tasks such as RNA-seq analysis with appropriate safeguards, rather than blocking all subsequent operations.
Loop detection middleware intercepts conversation with "[LOOP DETECTED] You are repeating the same tool calls..." message, causing LLM request failure with 400 error: InvalidParameter: Invalid messages at 12: {'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls...'}. This prevents completion of legitimate batch operations in scientific workflows.
Initiate RNA-seq differential expression analysis requiring multiple sequential tool calls
Execute batch operations (quality control, alignment, quantification) on multiple samples
Observe workflow interruption when loop detection triggers after repeated bash tool calls
Note the 400 error caused by loop detection middleware injecting error message as user content
logs/langgraph.log
openai.BadRequestError: Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-9cc7bf0c-332e-9cb9-ac4e-2b8bb58989f9', 'request_id': '9cc7bf0c-332e-9cb9-ac4e-2b8bb58989f9'}
�[2m2026-04-24T08:38:17.830628Z�[0m [�[32m�[1minfo �[0m] �[1m1 change detected �[0m [�[0m�[1m�[34mwatchfiles.main�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mthread_name�[0m=�[35mMainThread�[0m
�[2m2026-04-24T08:38:17.864915Z�[0m [�[32m�[1minfo �[0m] �[1mHTTP Request: POST https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions "HTTP/1.1 400 Bad Request"�[0m [�[0m�[1m�[34mhttpx�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m
�[2m2026-04-24T08:38:17.866061Z�[0m [�[33m�[1mwarning �[0m] �[1mLLM call failed after 1 attempt(s): Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-a8bc8c50-dc51-9db3-bf26-d7fd020d261b', 'request_id': 'a8bc8c50-dc51-9db3-bf26-d7fd020d261b'}�[0m [�[0m�[1m�[34mdeerflow.agents.middlewares.llm_error_handling_middleware�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m
Traceback (most recent call last):
File "/app/backend/packages/harness/deerflow/agents/middlewares/llm_error_handling_middleware.py", line 275, in awrap_model_call
response = await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 257, in inner_handler
inner_result = await inner(req, handler)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/middleware/todo.py", line 228, in awrap_model_call
return await handler(request.override(system_message=new_system_message))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 1156, in _execute_model_async
output = await model_.ainvoke(messages)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/runnables/base.py", line 5708, in ainvoke
return await self.bound.ainvoke(
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 477, in ainvoke
llm_result = await self.agenerate_prompt(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1196, in agenerate_prompt
return await self.agenerate(
^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1154, in agenerate
raise exceptions[0]
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1380, in _agenerate_with_cache
async for chunk in self._astream(messages, stop=stop, **kwargs):
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 3029, in _astream
async for chunk in super()._astream(*args, **kwargs):
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 1545, in _astream
response = await self.async_client.create(**payload)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/openai/resources/chat/completions/completions.py", line 2678, in create
return await self._post(
^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1797, in post
return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1597, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-a8bc8c50-dc51-9db3-bf26-d7fd020d261b', 'request_id': 'a8bc8c50-dc51-9db3-bf26-d7fd020d261b'}
�[2m2026-04-24T08:38:18.117544Z�[0m [�[32m�[1minfo �[0m] �[1mHTTP Request: POST https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions "HTTP/1.1 400 Bad Request"�[0m [�[0m�[1m�[34mhttpx�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m
�[2m2026-04-24T08:38:18.118628Z�[0m [�[33m�[1mwarning �[0m] �[1mLLM call failed after 1 attempt(s): Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-4020e9f2-f50e-9db4-9be8-095c993b871d', 'request_id': '4020e9f2-f50e-9db4-9be8-095c993b871d'}�[0m [�[0m�[1m�[34mdeerflow.agents.middlewares.llm_error_handling_middleware�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m
Traceback (most recent call last):
File "/app/backend/packages/harness/deerflow/agents/middlewares/llm_error_handling_middleware.py", line 275, in awrap_model_call
response = await handler(request)
^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 257, in inner_handler
inner_result = await inner(req, handler)
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/middleware/todo.py", line 228, in awrap_model_call
return await handler(request.override(system_message=new_system_message))
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 1156, in _execute_model_async
output = await model_.ainvoke(messages)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/runnables/base.py", line 5708, in ainvoke
return await self.bound.ainvoke(
^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 477, in ainvoke
llm_result = await self.agenerate_prompt(
^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1196, in agenerate_prompt
return await self.agenerate(
^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1154, in agenerate
raise exceptions[0]
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1380, in _agenerate_with_cache
async for chunk in self._astream(messages, stop=stop, **kwargs):
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 3029, in _astream
async for chunk in super()._astream(*args, **kwargs):
File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 1545, in _astream
response = await self.async_client.create(**payload)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/openai/resources/chat/completions/completions.py", line 2678, in create
return await self._post(
^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1797, in post
return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1597, in request
raise self._make_status_error_from_response(err.response) from None
openai.BadRequestError: Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-4020e9f2-f50e-9db4-9be8-095c993b871d', 'request_id': '4020e9f2-f50e-9db4-9be8-095c993b871d'}
logs/gateway.log
2026-04-24 08:43:52 - app.gateway.app - INFO - Configuration loaded successfully
2026-04-24 08:43:52 - app.gateway.app - INFO - Starting API Gateway on 0.0.0.0:8001
2026-04-24 08:43:52 - deerflow.runtime.stream_bridge.async_provider - INFO - Stream bridge initialised: memory (queue_maxsize=256)
2026-04-24 08:43:52 - deerflow.runtime.store.async_provider - INFO - Store: using AsyncSqliteStore (/app/backend/.deer-flow/checkpoints.db)
2026-04-24 08:43:52 - app.gateway.app - INFO - LangGraph runtime initialised
2026-04-24 08:43:52 - app.channels.manager - INFO - ChannelManager started (max_concurrency=5)
2026-04-24 08:43:52 - app.channels.service - INFO - ChannelService started with channels: []
2026-04-24 08:43:52 - app.gateway.app - INFO - Channel service started: {'service_running': True, 'channels': {'discord': {'enabled': False, 'running': False}, 'feishu': {'enabled': False, 'running': False}, 'slack': {'enabled': False, 'running': False}, 'telegram': {'enabled': False, 'running': False}, 'wechat': {'enabled': False, 'running': False}, 'wecom': {'enabled': False, 'running': False}}}
2026-04-24 08:43:52 - app.channels.manager - INFO - [Manager] dispatch loop started, waiting for inbound messages
INFO: Application startup complete.
WARNING: WatchFiles detected changes in 'packages/harness/deerflow/agents/middlewares/loop_detection_middleware.py'. Reloading...
INFO: Shutting down
INFO: Waiting for application shutdown.
2026-04-24 08:44:11 - app.channels.manager - INFO - ChannelManager stopped
2026-04-24 08:44:11 - app.channels.service - INFO - ChannelService stopped
2026-04-24 08:44:11 - app.gateway.app - INFO - Shutting down API Gateway
INFO: Application shutdown complete.
INFO: Finished server process [39]
INFO: Started server process [55]
INFO: Waiting for application startup.
Problem summary
Loop detection middleware incorrectly flags legitimate scientific workflows as infinite loops, blocking continuation of complex analyses after detecting repeated tool calls. This impacts RNA-seq analysis requiring multiple sequential operations.
Expected behavior
Loop detection should distinguish between actual infinite loops and legitimate scientific workflows requiring repeated operations. System should allow continued execution of batch processing tasks such as RNA-seq analysis with appropriate safeguards, rather than blocking all subsequent operations.
Actual behavior
Loop detection middleware intercepts conversation with "[LOOP DETECTED] You are repeating the same tool calls..." message, causing LLM request failure with 400 error: InvalidParameter: Invalid messages at 12: {'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls...'}. This prevents completion of legitimate batch operations in scientific workflows.
Operating system
Linux
Platform details
No response
Python version
Python 3.10.12
Node.js version
v22.21.0
pnpm version
No response
uv version
uv 0.8.9
How are you running DeerFlow?
Docker (make docker-dev)
Reproduction steps
Initiate RNA-seq differential expression analysis requiring multiple sequential tool calls
Execute batch operations (quality control, alignment, quantification) on multiple samples
Observe workflow interruption when loop detection triggers after repeated bash tool calls
Note the 400 error caused by loop detection middleware injecting error message as user content
rna-seq-differential-analysis.zip
Relevant logs
logs/langgraph.log openai.BadRequestError: Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-9cc7bf0c-332e-9cb9-ac4e-2b8bb58989f9', 'request_id': '9cc7bf0c-332e-9cb9-ac4e-2b8bb58989f9'} �[2m2026-04-24T08:38:17.830628Z�[0m [�[32m�[1minfo �[0m] �[1m1 change detected �[0m [�[0m�[1m�[34mwatchfiles.main�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mthread_name�[0m=�[35mMainThread�[0m �[2m2026-04-24T08:38:17.864915Z�[0m [�[32m�[1minfo �[0m] �[1mHTTP Request: POST https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions "HTTP/1.1 400 Bad Request"�[0m [�[0m�[1m�[34mhttpx�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m �[2m2026-04-24T08:38:17.866061Z�[0m [�[33m�[1mwarning �[0m] �[1mLLM call failed after 1 attempt(s): Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-a8bc8c50-dc51-9db3-bf26-d7fd020d261b', 'request_id': 'a8bc8c50-dc51-9db3-bf26-d7fd020d261b'}�[0m [�[0m�[1m�[34mdeerflow.agents.middlewares.llm_error_handling_middleware�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m Traceback (most recent call last): File "/app/backend/packages/harness/deerflow/agents/middlewares/llm_error_handling_middleware.py", line 275, in awrap_model_call response = await handler(request) ^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 257, in inner_handler inner_result = await inner(req, handler) ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/middleware/todo.py", line 228, in awrap_model_call return await handler(request.override(system_message=new_system_message)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 1156, in _execute_model_async output = await model_.ainvoke(messages) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/runnables/base.py", line 5708, in ainvoke return await self.bound.ainvoke( ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 477, in ainvoke llm_result = await self.agenerate_prompt( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1196, in agenerate_prompt return await self.agenerate( ^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1154, in agenerate raise exceptions[0] File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1380, in _agenerate_with_cache async for chunk in self._astream(messages, stop=stop, **kwargs): File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 3029, in _astream async for chunk in super()._astream(*args, **kwargs): File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 1545, in _astream response = await self.async_client.create(**payload) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/openai/resources/chat/completions/completions.py", line 2678, in create return await self._post( ^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1797, in post return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1597, in request raise self._make_status_error_from_response(err.response) from None openai.BadRequestError: Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-a8bc8c50-dc51-9db3-bf26-d7fd020d261b', 'request_id': 'a8bc8c50-dc51-9db3-bf26-d7fd020d261b'} �[2m2026-04-24T08:38:18.117544Z�[0m [�[32m�[1minfo �[0m] �[1mHTTP Request: POST https://dashscope.aliyuncs.com/compatible-mode/v1/chat/completions "HTTP/1.1 400 Bad Request"�[0m [�[0m�[1m�[34mhttpx�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m �[2m2026-04-24T08:38:18.118628Z�[0m [�[33m�[1mwarning �[0m] �[1mLLM call failed after 1 attempt(s): Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-4020e9f2-f50e-9db4-9be8-095c993b871d', 'request_id': '4020e9f2-f50e-9db4-9be8-095c993b871d'}�[0m [�[0m�[1m�[34mdeerflow.agents.middlewares.llm_error_handling_middleware�[0m]�[0m �[36mapi_variant�[0m=�[35mlocal_dev�[0m �[36massistant_id�[0m=�[35mbee7d354-5df5-5f26-a978-10ea053f620d�[0m �[36mgraph_id�[0m=�[35mlead_agent�[0m �[36mlanggraph_api_version�[0m=�[35m0.7.65�[0m �[36mlanggraph_node�[0m=�[35mmodel�[0m �[36mrequest_id�[0m=�[35m9a3d772c-ab23-4782-88ea-7c3d73f2f1c0�[0m �[36mrun_attempt�[0m=�[35m1�[0m �[36mrun_id�[0m=�[35m019dbe9b-9ba0-72c0-bdd5-73e68fe77561�[0m �[36mthread_id�[0m=�[35m765bc7c1-00ba-44a0-86fa-a371d4406bd9�[0m �[36mthread_name�[0m=�[35mMainThread�[0m Traceback (most recent call last): File "/app/backend/packages/harness/deerflow/agents/middlewares/llm_error_handling_middleware.py", line 275, in awrap_model_call response = await handler(request) ^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 257, in inner_handler inner_result = await inner(req, handler) ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/middleware/todo.py", line 228, in awrap_model_call return await handler(request.override(system_message=new_system_message)) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain/agents/factory.py", line 1156, in _execute_model_async output = await model_.ainvoke(messages) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/runnables/base.py", line 5708, in ainvoke return await self.bound.ainvoke( ^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 477, in ainvoke llm_result = await self.agenerate_prompt( ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1196, in agenerate_prompt return await self.agenerate( ^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1154, in agenerate raise exceptions[0] File "/app/backend/.venv/lib/python3.12/site-packages/langchain_core/language_models/chat_models.py", line 1380, in _agenerate_with_cache async for chunk in self._astream(messages, stop=stop, **kwargs): File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 3029, in _astream async for chunk in super()._astream(*args, **kwargs): File "/app/backend/.venv/lib/python3.12/site-packages/langchain_openai/chat_models/base.py", line 1545, in _astream response = await self.async_client.create(**payload) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/openai/resources/chat/completions/completions.py", line 2678, in create return await self._post( ^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1797, in post return await self.request(cast_to, opts, stream=stream, stream_cls=stream_cls) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/app/backend/.venv/lib/python3.12/site-packages/openai/_base_client.py", line 1597, in request raise self._make_status_error_from_response(err.response) from None openai.BadRequestError: Error code: 400 - {'error': {'message': "<400> InternalError.Algo.InvalidParameter: Error message is: Invalid messages at 12:\n{'role': 'user', 'content': '[LOOP DETECTED] You are repeating the same tool calls. Stop calling tools and produce your final answer now. If you cannot complete the task, summarize what you accomplished so far.<｜end▁of▁sentence｜><｜end▁of▁sentence｜><｜end▁of▁sentence｜>'}", 'type': 'invalid_request_error', 'param': None, 'code': 'invalid_parameter_error'}, 'id': 'chatcmpl-4020e9f2-f50e-9db4-9be8-095c993b871d', 'request_id': '4020e9f2-f50e-9db4-9be8-095c993b871d'} logs/gateway.log 2026-04-24 08:43:52 - app.gateway.app - INFO - Configuration loaded successfully 2026-04-24 08:43:52 - app.gateway.app - INFO - Starting API Gateway on 0.0.0.0:8001 2026-04-24 08:43:52 - deerflow.runtime.stream_bridge.async_provider - INFO - Stream bridge initialised: memory (queue_maxsize=256) 2026-04-24 08:43:52 - deerflow.runtime.store.async_provider - INFO - Store: using AsyncSqliteStore (/app/backend/.deer-flow/checkpoints.db) 2026-04-24 08:43:52 - app.gateway.app - INFO - LangGraph runtime initialised 2026-04-24 08:43:52 - app.channels.manager - INFO - ChannelManager started (max_concurrency=5) 2026-04-24 08:43:52 - app.channels.service - INFO - ChannelService started with channels: [] 2026-04-24 08:43:52 - app.gateway.app - INFO - Channel service started: {'service_running': True, 'channels': {'discord': {'enabled': False, 'running': False}, 'feishu': {'enabled': False, 'running': False}, 'slack': {'enabled': False, 'running': False}, 'telegram': {'enabled': False, 'running': False}, 'wechat': {'enabled': False, 'running': False}, 'wecom': {'enabled': False, 'running': False}}} 2026-04-24 08:43:52 - app.channels.manager - INFO - [Manager] dispatch loop started, waiting for inbound messages INFO: Application startup complete. WARNING: WatchFiles detected changes in 'packages/harness/deerflow/agents/middlewares/loop_detection_middleware.py'. Reloading... INFO: Shutting down INFO: Waiting for application shutdown. 2026-04-24 08:44:11 - app.channels.manager - INFO - ChannelManager stopped 2026-04-24 08:44:11 - app.channels.service - INFO - ChannelService stopped 2026-04-24 08:44:11 - app.gateway.app - INFO - Shutting down API Gateway INFO: Application shutdown complete. INFO: Finished server process [39] INFO: Started server process [55] INFO: Waiting for application startup.Git state
30d619d (HEAD -> g80140-dev, upstream/main, local/main, main) feat(subagents): support per-subagent skill loading and custom subagent types (#2253
Additional context
No response