Returning `Command` with `goto` attribute in a graph with >= 3 total graph depth results in `langgraph.errors.ParentCommand` error

[egebilecen]

Checked other resources

• [x] This is a bug, not a usage question. For questions, please use the LangChain Forum (https://forum.langchain.com/).
• [x] I added a clear and detailed title that summarizes the issue.
• [x] I read what a minimal reproducible example is (https://stackoverflow.com/help/minimal-reproducible-example).
• [x] I included a self-contained, minimal example that demonstrates the issue INCLUDING all the relevant imports. The code run AS IS to reproduce the issue.

Example Code

from langchain_core.messages import BaseMessage
from langchain_core.runnables import RunnableConfig
from langgraph.checkpoint.memory import MemorySaver
from langgraph.graph import StateGraph, START
from langgraph.types import Command
from langgraph.graph.message import add_messages
from typing import Annotated, TypedDict, Sequence

class State(TypedDict):
    messages: Annotated[Sequence[BaseMessage], add_messages]

def sub_sub_child(state):
    return Command(
        graph=Command.PARENT,
        goto="sub_child_3",
        update={"messages": state["messages"] + ["sub_sub_child executed, going to sub_child_3"]},
    )

sub_sub_builder = StateGraph(State)
sub_sub_builder.add_node("sub_sub_child", sub_sub_child)
sub_sub_builder.add_edge(START, "sub_sub_child")
sub_sub_graph = sub_sub_builder.compile(name="sub_sub_graph", checkpointer=True)

def sub_child_1(state):
    return {"messages": state["messages"] + ["sub_child_1 executed"]}

def sub_child_3(state):
    return {"messages": state["messages"] + ["sub_child_3 executed"]}

sub_builder = StateGraph(State)
sub_builder.add_node("sub_child_1", sub_child_1)
sub_builder.add_node("sub_child_2", sub_sub_graph)
sub_builder.add_node("sub_child_3", sub_child_3)
sub_builder.add_edge(START, "sub_child_1")
sub_builder.add_edge("sub_child_1", "sub_child_2")
sub_graph = sub_builder.compile(name="sub_graph", checkpointer=True)

def child_1(state):
    return {"messages": state["messages"] + ["child_1 executed"]}

builder = StateGraph(State)
builder.add_node("child_1", child_1)
builder.add_node("child_2", sub_graph)
builder.add_edge(START, "child_1")
builder.add_edge("child_1", "child_2")
graph = builder.compile(name="main_graph", checkpointer=MemorySaver())

result = graph.invoke({"messages":[]}, RunnableConfig(recursion_limit=100, configurable={"thread_id": 1}))

Error Message and Stack Trace (if applicable)

langgraph.errors.ParentCommand: Command(graph='child_2|sub_child_2', update={'messages': [HumanMessage(content='child_1 executed', additional_kwargs={}, response_metadata={}, id='0d913850-1062-4395-a94b-1edccb9fc3e8'), HumanMessage(content='sub_child_1 executed', additional_kwargs={}, response_metadata={}, id='bce46bfc-5040-40e7-a080-04c735edcabe'), 'sub_sub_child executed, going to sub_child_3']}, goto='sub_child_3')

Description

I am working on a graph which has a sub-graph as one of its nodes. This sub-graph also has a sub-graph in it. So essentially, we have a parent graph (called main_graph), which has one sub-graph as its node (called sub_graph). This sub_graph also has a sub-graph as its node (called sub_sub_graph). Returning a Command from the sub_sub_graph's one node that targets its parent (which should be sub_graph) results in an error.

System Info

System Information

OS: Linux OS Version: #1 SMP PREEMPT_DYNAMIC Wed Jan 15 08:02:15 EST 2025 Python Version: 3.12.0 (main, Aug 26 2024, 03:18:57) [GCC 8.5.0 20210514 (Red Hat 8.5.0-4)]

Package Information

langchain_core: 0.3.76 langchain: 0.3.27 langsmith: 0.4.4 langchain_openai: 0.3.27 langchain_text_splitters: 0.3.11 langgraph_sdk: 0.2.0 langgraph_supervisor: 0.0.29

Optional packages not installed

langserve

Other Dependencies

async-timeout<5.0.0,>=4.0.0;: Installed. No version info available. httpx: 0.28.1 httpx>=0.25.2: Installed. No version info available. jsonpatch<2.0,>=1.33: Installed. No version info available. langchain-anthropic;: Installed. No version info available. langchain-aws;: Installed. No version info available. langchain-azure-ai;: Installed. No version info available. langchain-cohere;: Installed. No version info available. langchain-community;: Installed. No version info available. langchain-core<1.0.0,>=0.3.66: Installed. No version info available. langchain-core<1.0.0,>=0.3.72: Installed. No version info available. langchain-core<2.0.0,>=0.3.75: Installed. No version info available. langchain-core>=0.3.40: Installed. No version info available. langchain-deepseek;: Installed. No version info available. langchain-fireworks;: Installed. No version info available. langchain-google-genai;: Installed. No version info available. langchain-google-vertexai;: Installed. No version info available. langchain-groq;: Installed. No version info available. langchain-huggingface;: Installed. No version info available. langchain-mistralai;: Installed. No version info available. langchain-ollama;: Installed. No version info available. langchain-openai;: Installed. No version info available. langchain-perplexity;: Installed. No version info available. langchain-text-splitters<1.0.0,>=0.3.9: Installed. No version info available. langchain-together;: Installed. No version info available. langchain-xai;: Installed. No version info available. langgraph<0.7.0,>=0.6.0: Installed. No version info available. langsmith-pyo3: Installed. No version info available. langsmith>=0.1.17: Installed. No version info available. langsmith>=0.3.45: Installed. No version info available. openai-agents: Installed. No version info available. openai<2.0.0,>=1.86.0: Installed. No version info available. opentelemetry-api: Installed. No version info available. opentelemetry-exporter-otlp-proto-http: Installed. No version info available. opentelemetry-sdk: Installed. No version info available. orjson: 3.10.18 orjson>=3.10.1: Installed. No version info available. packaging: 24.2 packaging>=23.2: Installed. No version info available. pydantic: 2.10.6 pydantic<3.0.0,>=2.7.4: Installed. No version info available. pydantic>=2.7.4: Installed. No version info available. pytest: Installed. No version info available. PyYAML>=5.3: Installed. No version info available. requests: 2.32.4 requests-toolbelt: 1.0.0 requests<3,>=2: Installed. No version info available. rich: 14.0.0 SQLAlchemy<3,>=1.4: Installed. No version info available. tenacity!=8.4.0,<10.0.0,>=8.1.0: Installed. No version info available. tiktoken<1,>=0.7: Installed. No version info available. typing-extensions>=4.7: Installed. No version info available. zstandard: 0.23.0

[le-codeur-rapide]

Hello everyone ! I was looking at this issue also, it looks like there might be an issue with how the checkpoint_ns is handle when ParentCommand is raised:

I found that the issue comes from this part of the code when the ParentCommand is raised:

# libs/langgraph/langgraph/pregel/_retry.py line 43
except ParentCommand as exc:
    ns: str = config[CONF][CONFIG_KEY_CHECKPOINT_NS]
    cmd = exc.args[0]
    if cmd.graph in (ns, task.name):
        # this command is for the current graph, handle it
        for w in task.writers:
            w.invoke(cmd, config)
        break

Here, with the 3-depth agent setup of the code sample of @egebilecen , we have cmd.graph = "child_2|sub_child_2" and ns = 'child_2:110740f9-a446-f6d4-d529-c2895838dedd|sub_child_2:360206dd-8b3b-bf9a-d24b-e716509f4a98', so the following tasks won't be executed.

Maybe we should reapply the recast_checkpoint_ns function at some point during the ParentCommand exception handeling ?

What do you think ?

PS: When we don't have 3 layers of agents, (for example in the code sample of the issue, replace graph.invoke by sub_graph.invoke, we have cmd.graph = sub_child_2 and ns = 'sub_child_2:f7037092-7563-7b36-5b4b-993f905bab47' so cmd.graph in ns == True This explains why the ParentCommand works if there are only a parent graph and a sub graph

[egebilecen]

Thanks for the quick fix. Do we have any ETA for merge? It is needed for one of the projects in my company.