Langchain Langgraph Basics

LangChain LangGraph Basics (Python)

Overview

A conditional edge whose router returns a string that is not in path_map halts the graph without reaching END. No exception. No log line. The invocation just returns whatever state existed at the halt point — pain-catalog entry P56, and the single most common reason a newly wired StateGraph "almost works." The sibling pain: Command(update={"messages": [msg]}) wipes the prior message history because messages was declared as a plain list[AnyMessage] instead of Annotated[list[AnyMessage], add_messages] — the reducer is what turns update into "append" instead of "replace" (P18).

Two more gotchas this skill defuses:

• P55 — GraphRecursionError: Recursion limit of 25 reached fires on graphs that never loop, because recursion_limit counts supersteps (one step per synchronous batch of node executions), not loop iterations. A planner
  • executor + validator + summarizer can hit 25 without any cycle.
• P20 — Upgrading langgraph silently reads old PostgresSaver checkpoints as empty state. Checkpoint schemas evolve; PostgresSaver.setup() must be rerun after every version bump before production traffic.

This skill walks through a minimal StateGraph end to end: a TypedDict state with reducers on every list field, node functions that return partial-state dicts, edges and defensive conditional edges with END as a fallback in path_map, compilation with a checkpointer, recursion_limit sizing, and invocation with an explicit thread_id. Pin: langgraph 1.0.x, langchain-core 1.0.x. Pain-catalog anchors: P16, P18, P20, P55, P56.

Prerequisites

• Python 3.10+
• pip install langgraph>=1.0,<2.0 langchain-core>=1.0,<2.0
• A chat model (see langchain-model-inference), or a pure-logic graph with no LLM
• For persistence beyond a single process: pip install langgraph-checkpoint-postgres and a Postgres 14+ instance

Instructions

Step 1 — Define state as a TypedDict with reducers on list fields

Every list-shaped field in state needs a reducer. Without one, Command(update=...) and node returns replace the field. The message-history reducer lives in langgraph.graph.message:

from typing import Annotated, TypedDict
from langchain_core.messages import AnyMessage
from langgraph.graph.message import add_messages
import operator

class AgentState(TypedDict):
    # Reducer "add_messages" appends + dedupes by message id (P18)
    messages: Annotated[list[AnyMessage], add_messages]

    # Plain list field also needs a reducer — use operator.add to concat
    scratchpad: Annotated[list[str], operator.add]

    # Scalars don't need a reducer; update replaces them
    step_count: int
    done: bool

If you forget the reducer on messages, a resume with Command(update={"messages": [new_msg]}) will overwrite the entire prior history. Validate reducers are in place with graph.get_graph().draw_mermaid() — annotated fields render with their reducer name.

See State Reducers for the built-in list (add_messages, operator.add, max, min) and how to write a custom merger for non-trivial merge logic.

Step 2 — Write nodes as functions that return partial-state dicts

A node takes the full state and returns only the keys it wants to update. The reducer handles merge:

def plan(state: AgentState) -> dict:
    # Returning a dict means "update these fields"
    return {
        "messages": [("assistant", "Plan: step 1, step 2, step 3")],
        "scratchpad": ["planned_at_step_1"],
        "step_count": state["step_count"] + 1,
    }

def execute(state: AgentState) -> dict:
    return {
        "messages": [("assistant", f"Executed {state['step_count']} steps")],
        "done": state["step_count"] >= 3,
    }

Nodes must be deterministic on their inputs — LangGraph re-runs them during time-travel replay, and a side-effecting node (DB write without idempotency key) will double-fire. Push side effects to the checkpointer boundary or tool calls.

Step 3 — Wire edges and conditional edges defensively

from typing import Literal
from langgraph.graph import StateGraph, START, END

# Router MUST return a value in the path_map keyset (P56)
def should_continue(state: AgentState) -> Literal["execute", "end"]:
    if state["done"] or state["step_count"] >= 10:
        return "end"
    return "execute"

builder = StateGraph(AgentState)
builder.add_node("plan", plan)
builder.add_node("execute", execute)

builder.add_edge(START, "plan")

# path_map ALWAYS includes END as a fallback — if the router returns anything
# else, the graph reaches END instead of halting silently (P56)
builder.add_conditional_edges(
    "plan",
    should_continue,
    path_map={"execute": "execute", "end": END},
)
builder.add_edge("execute", "plan")  # loop back to plan

The Literal return annotation on should_continue is a static guard — mypy catches typos before runtime. path_map={"execute": "execute", "end": END} is the spelled-out form; the compact form path_map=["execute", END] also works when router return values match node names directly.

See Conditional Edges for all four add_conditional_edges signatures, the path vs path_map distinction, and a pytest pattern that asserts every router return value hits a known route.

Step 4 — Compile with a checkpointer

from langgraph.checkpoint.memory import MemorySaver

# MemorySaver is in-process — use PostgresSaver in production (P20)
checkpointer = MemorySaver()
graph = builder.compile(checkpointer=checkpointer)

For production, swap to langgraph.checkpoint.postgres.PostgresSaver. After every langgraph version bump, run PostgresSaver.setup() in staging before prod traffic — the schema evolves and old rows are silently read as empty state.

Step 5 — Pick recursion_limit for the graph's superstep count

recursion_limit defaults to 25. It is not a loop counter; it counts total supersteps, and a superstep is one synchronous round of node executions (parallel branches in the same step count as one). Typical shapes:

config = {
    "configurable": {"thread_id": "user-42"},  # required for checkpointing (P16)
    "recursion_limit": 30,
}
result = graph.invoke({"messages": [], "scratchpad": [], "step_count": 0, "done": False}, config)

If you hit GraphRecursionError on a graph that clearly isn't looping (P55), add print(state["step_count"]) at the entry of each node to see the actual superstep count, then either raise the limit or restructure with a subgraph so each subgraph gets its own budget.

See Recursion Limits for the full derivation and a diagnostic script that traces superstep count at runtime.

Step 6 — Invoke with thread_id in config["configurable"]

Every invocation against a checkpointer-backed graph needs a thread_id in config["configurable"]. Without it, each call gets a fresh state with no warning (P16). Enforce it at your application boundary:

def run_agent(user_id: str, user_message: str) -> dict:
    config = {
        "configurable": {"thread_id": user_id},
        "recursion_limit": 30,
    }
    assert config["configurable"].get("thread_id"), "thread_id required"
    return graph.invoke(
        {"messages": [("user", user_message)], "scratchpad": [], "step_count": 0, "done": False},
        config,
    )

See First Graph Walkthrough for a line-by-line annotation of a minimal 3-node graph that demonstrates typed state, a reducer, and a conditional edge to END.

Decision Tree

Is your field a list you want to append?
  -> Annotate with add_messages (for messages) or operator.add (for plain lists)

Is your router adding a new output string?
  -> Add that string to path_map BEFORE deploying; include END as a fallback

Hitting GraphRecursionError on a non-looping graph?
  -> supersteps != iterations; raise recursion_limit to 50 or split into subgraphs

Upgraded langgraph minor version?
  -> Re-run PostgresSaver.setup() in staging before routing prod traffic

Multi-turn agent forgets between calls?
  -> thread_id missing from config["configurable"] — enforce at boundary

Output

• TypedDict state with reducer annotations on every list field
• Node functions returning partial-state dicts, free of hidden side effects
• Conditional edges with Literal-typed routers and END in every path_map
• Compiled graph with a checkpointer matched to the deployment shape
• recursion_limit sized from the superstep count table, not the default 25
• Invocations with explicit thread_id validated at the app boundary

Error Handling

Examples

Building a minimal linear graph (no loop)

Three nodes (plan → execute → summarize), no conditionals, one reducer on messages. Runs in 4 supersteps (START counts as one). Safe at default recursion_limit=25, but set it to 10 explicitly so readers see the budget.

See First Graph Walkthrough for the line-by-line annotation and the graph.get_graph().draw_mermaid() output.

Adding a conditional loop with a bounded retry budget

A validator node that either completes ("end" → END) or retries ("retry" → back to executor), bounded by step_count >= 3. The router is Literal["retry", "end"] and path_map maps both. Caps at 7 supersteps, so recursion_limit=15 is plenty of headroom.

See Conditional Edges for the full example and the pytest that iterates every Literal branch.

Diagnosing a mystery GraphRecursionError

A planner that fans out to 4 parallel executors, then a validator, then a summarizer. Looks linear in the mermaid diagram, hits GraphRecursionError: Recursion limit of 25 reached on 10% of runs. Cause: each parallel executor counts as its own superstep branch when the merge node is conditional. Fix: raise to 50 or wrap the fan-out in a subgraph.

See Recursion Limits for the diagnostic script and the subgraph refactor.

Resources

• LangGraph: Low-level concepts (StateGraph, reducers, nodes, edges)
• LangGraph: Persistence and checkpointing
• LangGraph: add_messages reducer
• LangGraph: add_conditional_edges API
• LangGraph: Recursion limit and supersteps
• LangChain 1.0 release notes
• Pack pain catalog: docs/pain-catalog.md (entries P16, P17, P18, P20, P55, P56)

Source: flight505/skill-forge — distributed by TomeVault.