After SummarizationMiddleware is triggered, agent can randomly triggers 400 - No tool call found #34282
Description
Checked other resources
- This is a bug, not a usage question.
- I added a clear and descriptive title that summarizes this issue.
- I used the GitHub search to find a similar question and didn't find it.
- I am sure that this is a bug in LangChain rather than my code.
- The bug is not resolved by updating to the latest stable version of LangChain (or the specific integration package).
- This is not related to the langchain-community package.
- I posted a self-contained, minimal, reproducible example. A maintainer can copy it and run it AS IS.
Package (Required)
- langchain
- langchain-openai
- langchain-anthropic
- langchain-classic
- langchain-core
- langchain-cli
- langchain-model-profiles
- langchain-tests
- langchain-text-splitters
- langchain-chroma
- langchain-deepseek
- langchain-exa
- langchain-fireworks
- langchain-groq
- langchain-huggingface
- langchain-mistralai
- langchain-nomic
- langchain-ollama
- langchain-perplexity
- langchain-prompty
- langchain-qdrant
- langchain-xai
- Other / not sure / general
Example Code (Python)
-Error Message and Stack Trace (if applicable)
BadRequestError("Error code: 400 - {'error': {'message': 'No tool call found for function call output with call_id call_isCU6SdR5T4GBrTwrixhapmP.', 'type': 'invalid_request_error', 'param': 'input', 'code': None}}")Traceback (most recent call last): File "/my-proj/.venv/lib/python3.14/site-packages/langgraph/pregel/main.py", line 2643, in stream for _ in runner.tick( ~~~~~~~~~~~^ [t for t in loop.tasks.values() if not t.writes], ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ...<2 lines>... schedule_task=loop.accept_push, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ): ^ File "/my-proj/.venv/lib/python3.14/site-packages/langgraph/pregel/_runner.py", line 167, in tick run_with_retry( ~~~~~~~~~~~~~~^ t, ^^ ...<10 lines>... }, ^^ ) ^ File "/my-proj/.venv/lib/python3.14/site-packages/langgraph/pregel/_retry.py", line 42, in run_with_retry return task.proc.invoke(task.input, config) ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^ File "/my-proj/.venv/lib/python3.14/site-packages/langgraph/_internal/_runnable.py", line 656, in invoke input = context.run(step.invoke, input, config, **kwargs) File "/my-proj/.venv/lib/python3.14/site-packages/langgraph/_internal/_runnable.py", line 400, in invoke ret = self.func(*args, **kwargs) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 1132, in model_node response = wrap_model_call_handler(request, _execute_model_sync) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 192, in final_normalized final_result = result(request, handler) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 176, in composed outer_result = outer(request, inner_handler) File "/my-proj/common/utils.py", line 149, in wrap_model_call return handler(request) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 172, in inner_handler inner_result = inner(req, handler) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 176, in composed outer_result = outer(request, inner_handler) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/middleware/todo.py", line 207, in wrap_model_call return handler(request.override(system_message=new_system_message)) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 172, in inner_handler inner_result = inner(req, handler) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 176, in composed outer_result = outer(request, inner_handler) File "/my-proj/common/utils.py", line 149, in wrap_model_call return handler(request) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 172, in inner_handler inner_result = inner(req, handler) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 176, in composed outer_result = outer(request, inner_handler) File "/my-proj/.venv/lib/python3.14/site-packages/deepagents/middleware/filesystem.py", line 763, in wrap_model_call return handler(request) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 172, in inner_handler inner_result = inner(req, handler) File "/my-proj/.venv/lib/python3.14/site-packages/langchain_anthropic/middleware/prompt_caching.py", line 112, in wrap_model_call return handler(request) File "/my-proj/.venv/lib/python3.14/site-packages/langchain/agents/factory.py", line 1102, in _execute_model_sync output = model_.invoke(messages) File "/my-proj/.venv/lib/python3.14/site-packages/langchain_core/runnables/base.py", line 5534, in invoke return self.bound.invoke( ~~~~~~~~~~~~~~~~~^ input, ^^^^^^ self._merge_configs(config), ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ **{**self.kwargs, **kwargs}, ^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ) ^ File "/my-proj/.venv/lib/python3.14/site-packages/langchain_core/language_models/chat_models.py", line 398, in invoke self.generate_prompt( ~~~~~~~~~~~~~~~~~~~~^ [self._convert_input(input)], ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ...<6 lines>... **kwargs, ^^^^^^^^^ ).generations[0][0], ^ File "/my-proj/.venv/lib/python3.14/site-packages/langchain_core/language_models/chat_models.py", line 1117, in generate_prompt return self.generate(prompt_messages, stop=stop, callbacks=callbacks, **kwargs) ~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/my-proj/.venv/lib/python3.14/site-packages/langchain_core/language_models/chat_models.py", line 927, in generate self._generate_with_cache( ~~~~~~~~~~~~~~~~~~~~~~~~~^ m, ^^ ...<2 lines>... **kwargs, ^^^^^^^^^ ) ^ File "/my-proj/.venv/lib/python3.14/site-packages/langchain_core/language_models/chat_models.py", line 1221, in _generate_with_cache result = self._generate( messages, stop=stop, run_manager=run_manager, **kwargs ) File "/my-proj/.venv/lib/python3.14/site-packages/langchain_openai/chat_models/base.py", line 1356, in _generate raise e File "/my-proj/.venv/lib/python3.14/site-packages/langchain_openai/chat_models/base.py", line 1338, in _generate raw_response = self.root_client.responses.with_raw_response.create( **payload ) File "/my-proj/.venv/lib/python3.14/site-packages/openai/_legacy_response.py", line 364, in wrapped return cast(LegacyAPIResponse[R], func(*args, **kwargs)) ~~~~^^^^^^^^^^^^^^^^^ File "/my-proj/.venv/lib/python3.14/site-packages/openai/resources/responses/responses.py", line 859, in create return self._post( ~~~~~~~~~~^ "/responses", ^^^^^^^^^^^^^ ...<40 lines>... stream_cls=Stream[ResponseStreamEvent], ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ) ^ File "/my-proj/.venv/lib/python3.14/site-packages/openai/_base_client.py", line 1259, in post return cast(ResponseT, self.request(cast_to, opts, stream=stream, stream_cls=stream_cls)) ~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/my-proj/.venv/lib/python3.14/site-packages/openai/_base_client.py", line 1047, in request raise self._make_status_error_from_response(err.response) from None openai.BadRequestError: Error code: 400 - {'error': {'message': 'No tool call found for function call output with call_id call_isCU6SdR5T4GBrTwrixhapmP.', 'type': 'invalid_request_error', 'param': 'input', 'code': None}} During task with name 'model' and id '3ba2f8f3-73ca-fb2f-e380-e8b6a9f289a9'Description
Using latest version of langchain.
using vanilla create_deep_agent() call, which contain the SummarisationMiddleware to keep 6 messages and requires 170k tokens to summarize.
Basically, summarisation should:
- Not summarise tools_call requests, in which the correspondent tool_call response wasn't yet received.
- If one of last messages to keep is a tool_call response, be sure that the correspondent tool_call request is also present in the messages (even if it's in the windows to summarise).