Build ShopBot: a LangGraph Support Agent With Memory, Tools and Refund Approval
Build a customer support agent in LangGraph module by module: state, tools, nodes, routing, rolling summaries and an interrupt that waits for refund approval.
This article is published in English.
Sources: Your First Real LangGraph Project: Building a Customer Support Agent.
This project pulls three LangGraph ideas into one agent: graph architecture, conversation memory and human-in-the-loop approval. The agent is a customer support bot called ShopBot for an imaginary e-commerce store. Customers can ask about orders, request refunds and escalate complaints. The bot remembers the conversation, looks up order data with real tools and stops for a person's approval before any refund goes through. The code is built up module by module with each piece explained, and prior LangGraph experience isn't required.
Plan the graph first
A common beginner mistake is opening an empty file and typing. LangGraph rewards drawing the graph first, on paper or in your head. On paper ShopBot is small: the agent step, the tool runner, the refund review and the summarizer make up the nodes, two branches decide the flow, a single interrupt() pauses for approval, and one trigger fires summarization. The code only transcribes that picture.
Setup
Make a file called shopbot.py, or open a Colab notebook. Install the dependencies:
pip install langgraph langchain langchain-openai langgraph-checkpoint-sqlite python-dotenv
Then create a .env file in the same folder to hold the key:
OPENAI_API_KEY=your-key-here
Module 1: imports and configuration
Every import for the whole file goes at the top, with nothing imported halfway down. Configuration lives here too.
# ============================================================
# MODULE 1: IMPORTS & CONFIGURATION
# ============================================================
import os
import sqlite3
from typing import Annotated, Literal
from datetime import datetime
from dotenv import load_dotenv
# LangChain - the AI layer
from langchain_openai import ChatOpenAI
from langchain_core.messages import (
HumanMessage,
AIMessage,
SystemMessage,
BaseMessage,
RemoveMessage,
)
from langchain_core.tools import tool
# LangGraph - the graph layer
from langgraph.graph import StateGraph, MessagesState, START, END
from langgraph.prebuilt import ToolNode
from langgraph.checkpoint.memory import MemorySaver
from langgraph.types import interrupt, Command
load_dotenv()
# ── The LLM ─────────────────────────────────────────────────
# temperature=0 means deterministic - the agent behaves
# consistently, which is what you want for a support bot.
llm = ChatOpenAI(model="gpt-4o-mini", temperature=0)
The model choice is gpt-4o-mini. It's capable enough for support work, cheap enough to call often and quick enough that users won't notice delay. A production system might move to gpt-4o for harder reasoning, but starting small is sensible.
Module 2: state
State is the agent's working memory, so everything that travels through the graph has to live in it. Decide what the agent must know at each stage.
# ============================================================
# MODULE 2: STATE
# ============================================================
class SupportState(MessagesState):
# MessagesState already gives us:
# messages: Annotated[list[BaseMessage], add_messages]
# We add three more fields for our specific needs:
# The running summary of the conversation (Part 2 pattern).
# Starts empty. Gets written by summarize_node when conversation gets long.
summary: str
# The name of the customer, extracted early in the conversation.
# Used to personalise every response. Starts empty.
customer_name: str
# Tracks the current ticket category, set by the agent.
# Helps the human reviewer understand context during escalation.
# Values: "order_inquiry" | "refund_request" | "complaint" | "general"
ticket_category: str
There are three extra fields beyond the message history. summary holds a compressed record of earlier turns: once a conversation runs past six messages, the agent condenses history into this field and deletes the old raw messages, so token costs stay flat. customer_name is a practical field in almost any support agent. With the name in state, the agent can greet people personally without each node re-reading the chat to find it. ticket_category is what makes conditional routing possible, since the agent sets it and a routing function reads it to decide whether the human approval path should trigger.
Module 3: tools
Tools are the agent's hands. Without them it can only talk, and with them it can look things up and take actions. ShopBot gets three, backed by a fake in-memory database:
# ============================================================
# MODULE 3: TOOLS
# ============================================================
# ── Fake Database ────────────────────────────────────────────
# In a real project, these would be database queries or API calls.
# For learning purposes, we use a simple Python dictionary.
ORDERS_DB = {
"ORD-001": {
"customer": "Alex",
"product": "Wireless Headphones",
"status": "Delivered",
"amount": 89.99,
"delivery_date": "2025-06-10",
},
"ORD-002": {
"customer": "Sam",
"product": "Phone Case",
"status": "In Transit",
"amount": 14.99,
"delivery_date": "Expected 2025-06-18",
},
"ORD-003": {
"customer": "Jordan",
"product": "Laptop Stand",
"status": "Processing",
"amount": 45.00,
"delivery_date": "Expected 2025-06-20",
},
}
@tool
def lookup_order(order_id: str) -> str:
"""Look up the details of a customer's order by order ID.
Use this when the customer provides an order number and wants
to know the status, product name, or delivery date of their order.
Args:
order_id: The order ID string, e.g. 'ORD-001'
Returns:
A formatted string with full order details, or an error message
if the order is not found.
"""
order = ORDERS_DB.get(order_id.upper())
if not order:
return f"No order found with ID '{order_id}'. Please double-check the order number."
return (
f"Order {order_id.upper()}: {order['product']} | "
f"Status: {order['status']} | "
f"Amount: ${order['amount']:.2f} | "
f"Delivery: {order['delivery_date']}"
)
@tool
def check_refund_eligibility(order_id: str) -> str:
"""Check whether an order is eligible for a refund.
Use this BEFORE processing any refund request. An order is eligible
for a refund only if its status is 'Delivered'. Orders Fthat are
'In Transit' or 'Processing' cannot be refunded yet.
Args:
order_id: The order ID string, e.g. 'ORD-001'
Returns:
A string stating whether the order is eligible and why.
"""
order = ORDERS_DB.get(order_id.upper())
if not order:
return f"Cannot check refund: order '{order_id}' not found."
if order["status"] == "Delivered":
return (
f"Order {order_id.upper()} IS eligible for a refund. "
f"Product: {order['product']}, Amount: ${order['amount']:.2f}. "
f"Proceed to refund processing."
)
else:
return (
f"Order {order_id.upper()} is NOT eligible for a refund yet. "
f"Current status: {order['status']}. Refunds are only available "
f"for delivered orders."
)
@tool
def process_refund(order_id: str, reason: str) -> str:
"""Process a refund for a delivered order.
IMPORTANT: This tool actually issues the refund. It should only be
called AFTER human approval has been obtained. Never call this tool
without prior confirmation.
Args:
order_id: The order ID to refund
reason: The customer's stated reason for the refund
Returns:
A confirmation string with the refund reference number.
"""
order = ORDERS_DB.get(order_id.upper())
if not order:
return f"Refund failed: order '{order_id}' not found."
# In a real system, this would hit your payments API.
refund_ref = f"REF-{order_id.upper()}-{datetime.now().strftime('%H%M%S')}"
return (
f"Refund APPROVED and PROCESSED. Reference: {refund_ref}. "
f"${order['amount']:.2f} will be returned to the original payment method "
f"within 3–5 business days. Reason logged: '{reason}'."
)
# ── Collect tools and bind to LLM ───────────────────────────
# All three tools in one list.
tools = [lookup_order, check_refund_eligibility, process_refund]
# llm_with_tools = the LLM that KNOWS about the tools and can decide to call them.
# This is what we use inside agent_node.
llm_with_tools = llm.bind_tools(tools)
# tool_node = the pre-built node that EXECUTES whatever tool the LLM chose.
# This is what we register in Module 6.
tool_node = ToolNode(tools)
Look closely at the docstring of process_refund, which says the tool really issues the refund and should be called only after human approval. The model reads docstrings and uses them when deciding what to do. Write them as if briefing a smart intern who doesn't yet know your business rules, because that's effectively what's happening.
Module 4: nodes
Nodes do the work. Each receives the state, does something and returns a dictionary of updates, nothing more. ShopBot has three nodes you write, plus the pre-built tool_node from the previous module. They include the system prompt, the main agent step, the refund review step with interrupt() and the summarizer:
# ============================================================
# MODULE 4: NODES
# ============================================================
# ── The System Prompt ────────────────────────────────────────
# Written once, used in every call to the LLM from agent_node.
# This is the personality and rulebook of your agent.
SYSTEM_PROMPT = """You are ShopBot, a friendly and professional customer support \
agent for an e-commerce store.
Your capabilities:
- Look up order details using the lookup_order tool
- Check if an order qualifies for a refund using check_refund_eligibility
- Process approved refunds using the process_refund tool
Your rules:
- Always greet the customer by name once you know it
- Always check refund eligibility BEFORE attempting to process a refund
- For refund requests, set ticket_category to "refund_request" in your reasoning
- Be empathetic, clear, and concise
- If you cannot help, offer to escalate to a human agent
Important: The process_refund tool requires prior human approval. Do not call it \
unless the conversation shows that a human has already approved the refund."""
# ── Node 1: agent_node ──────────────────────────────────────
def agent_node(state: SupportState) -> dict:
"""The brain of the operation. Reads state, calls the LLM, and decides
whether to use a tool, give a final answer, or do something else.
This node handles two cases:
1. Normal conversation - just call the LLM and respond
2. Long conversation - if a summary exists, prepend it so the LLM
has context without seeing all the raw messages
"""
# Part 2 pattern: check for an existing summary
summary = state.get("summary", "")
if summary:
# Build context: system prompt + compressed history + recent messages
system_with_summary = SystemMessage(
content=f"{SYSTEM_PROMPT}\n\nSummary of conversation so far:\n{summary}"
)
messages_to_send = [system_with_summary] + state["messages"]
else:
# No summary yet - full history is short enough to send as-is
system_msg = SystemMessage(content=SYSTEM_PROMPT)
messages_to_send = [system_msg] + state["messages"]
# Call the LLM. It sees tools and can choose to call one.
response = llm_with_tools.invoke(messages_to_send)
# Detect ticket category from the response for routing purposes.
# A smarter version would have the LLM explicitly set this -
# for now, we scan for keywords.
content_lower = response.content.lower() if response.content else ""
updates: dict = {"messages": [response]}
if "refund" in content_lower or (
hasattr(response, "tool_calls")
and any("refund" in str(tc).lower() for tc in (response.tool_calls or []))
):
updates["ticket_category"] = "refund_request"
return updates
# ── Node 2: review_refund ───────────────────────────────────
def review_refund(state: SupportState) -> dict:
"""The human approval gate. Pauses execution, shows the pending refund
details to a human agent, and waits for their decision.
This implements the Part 3 interrupt() pattern. Execution stops here
until someone calls graph.invoke(Command(resume=...), config).
Three outcomes the human can choose:
- "approve" → let the refund tool call proceed unchanged
- "reject" → cancel the refund, send a message to the customer
- "escalate" → hand the entire ticket to a human support agent
"""
last_message = state["messages"][-1]
# Find the refund-related tool call in the last AI message.
# We look for process_refund specifically - the "real action" tool.
refund_tool_call = None
if hasattr(last_message, "tool_calls"):
for tc in last_message.tool_calls:
if "refund" in tc["name"].lower():
refund_tool_call = tc
break
# Surface the context to the human reviewer via interrupt().
# Everything in this dict is what the human sees before deciding.
human_decision = interrupt({
"message": "⚠️ Refund approval required",
"customer_name": state.get("customer_name", "Unknown"),
"tool_being_called": refund_tool_call["name"] if refund_tool_call else "refund tool",
"arguments": refund_tool_call["args"] if refund_tool_call else {},
"conversation_summary": state.get("summary", "No summary yet"),
"options": ["approve", "reject", "escalate"],
})
# ── Handle the human's decision ─────────────────────────
if human_decision == "approve":
# Do nothing to state - let tool_node execute the tool call as-is
return {}
elif human_decision == "reject":
# Cancel the tool call. The LLM will see a ToolMessage explaining why,
# and generate a polite response to the customer.
from langchain_core.messages import ToolMessage
return {
"messages": [
ToolMessage(
content=(
"Refund request was reviewed and declined by our support team. "
"Please inform the customer politely and offer alternatives."
),
tool_call_id=refund_tool_call["id"] if refund_tool_call else "unknown",
)
]
}
elif human_decision == "escalate":
# Signal escalation - in a real system you'd open a ticket,
# ping Slack, or transfer to a live agent queue.
from langchain_core.messages import ToolMessage
return {
"messages": [
ToolMessage(
content=(
"This ticket has been escalated to a senior support agent. "
"Inform the customer that a human agent will contact them "
"within 2 business hours."
),
tool_call_id=refund_tool_call["id"] if refund_tool_call else "unknown",
)
]
}
# Fallback - treat as approve
return {}
# ── Node 3: summarize_node ──────────────────────────────────
def summarize_node(state: SupportState) -> dict:
"""Triggered when the conversation exceeds 6 messages. Compresses the
full message history into a short summary, then deletes old raw messages.
This is the rolling summary pattern from Part 2. The summary grows
richer turn by turn. Token costs stay nearly flat no matter how long
the conversation runs.
"""
existing_summary = state.get("summary", "")
if existing_summary:
# Extend the existing summary with new messages
summary_instruction = (
f"Current summary:\n{existing_summary}\n\n"
"Extend this summary with the new messages above. "
"Keep it under 5 sentences. Focus on: the customer's name, "
"their issue, any orders mentioned, and what actions were taken."
)
else:
# First time summarising
summary_instruction = (
"Summarise this customer support conversation in under 5 sentences. "
"Include: the customer's name (if mentioned), their issue, "
"any order numbers discussed, and what actions were taken so far."
)
messages = state["messages"] + [HumanMessage(content=summary_instruction)]
response = llm.invoke(messages) # Plain llm, no tools needed here
# Delete all but the 2 most recent messages.
# The summary now holds everything that was in the deleted messages.
messages_to_delete = [
RemoveMessage(id=m.id) for m in state["messages"][:-2]
]
return {
"summary": response.content,
"messages": messages_to_delete,
}
Module 5: edges and routing
Routing functions decide where to go next. They inspect state and return a string, and they never modify state, since that's the nodes' job. Two are needed:
# ============================================================
# MODULE 5: EDGES & ROUTING
# ============================================================
def route_after_agent(state: SupportState) -> Literal[
"review_refund", "tools", "summarize_node", "__end__"
]:
"""Called after agent_node runs. Decides what happens next.
Four possible routes:
1. The LLM wants to call process_refund → must go through human review first
2. The LLM wants to call any other tool → go directly to tool_node
3. The LLM gave a plain text answer AND the conversation is long → summarise
4. The LLM gave a plain text answer and conversation is short → we're done
"""
last_message = state["messages"][-1]
has_tool_calls = hasattr(last_message, "tool_calls") and bool(last_message.tool_calls)
if has_tool_calls:
# Check if ANY of the tool calls is the sensitive process_refund tool
tool_names = [tc["name"] for tc in last_message.tool_calls]
if "process_refund" in tool_names:
return "review_refund" # → Pause for human approval first
return "tools" # → Safe tool, run it directly
# No tool call - the LLM gave a plain response.
# Check if the conversation is long enough to need summarisation.
if len(state["messages"]) > 6:
return "summarize_node"
return "__end__" # → Conversation turn is complete
def route_after_review(state: SupportState) -> Literal["tools", "agent_node"]:
"""Called after review_refund runs (i.e., after the human has decided).
Two routes:
1. Human approved or escalated → run the tool (tool_node handles the call)
2. Human rejected → the review node already added a ToolMessage cancelling
the tool call, so skip tool_node and go back to agent_node to respond
"""
last_message = state["messages"][-1]
# If the last message is a ToolMessage, the review node cancelled the call.
# Go back to agent_node so it can generate a customer-facing response.
from langchain_core.messages import ToolMessage
if isinstance(last_message, ToolMessage):
return "agent_node"
# Otherwise, the review node returned {} (approved) - proceed to tools.
return "tools"
Module 6: assembling the graph
All the pieces now exist, and assembly always follows five steps: initialize, register nodes, set the entry point, add edges and compile.
# ============================================================
# MODULE 6: GRAPH ASSEMBLY
# ============================================================
# ── Step 1: Initialize ──────────────────────────────────────
graph_builder = StateGraph(SupportState)
# ── Step 2: Register All Nodes ──────────────────────────────
# Format: add_node("string_name", function)
# The string name is what you use in every edge definition below.
graph_builder.add_node("agent_node", agent_node)
graph_builder.add_node("tools", tool_node) # Pre-built from Module 3
graph_builder.add_node("review_refund", review_refund)
graph_builder.add_node("summarize_node", summarize_node)
# ── Step 3: Set Entry Point ─────────────────────────────────
# The first node that runs when a user sends a message.
graph_builder.add_edge(START, "agent_node")
# ── Step 4: Wire the Edges ──────────────────────────────────
# After agent_node: conditional - depends on what the LLM decided
graph_builder.add_conditional_edges(
"agent_node", # Source
route_after_agent, # Router function from Module 5
{
"review_refund": "review_refund", # Refund tool → human review first
"tools": "tools", # Other tools → run directly
"summarize_node": "summarize_node", # Long conversation → summarise
"__end__": END, # Plain answer → done
}
)
# After review_refund: conditional - depends on human's decision
graph_builder.add_conditional_edges(
"review_refund",
route_after_review,
{
"tools": "tools", # Approved → execute the tool
"agent_node": "agent_node", # Rejected → back to agent to respond
}
)
# After tools run: always go back to agent_node
# (the ReAct loop - agent sees tool result, decides what to do next)
graph_builder.add_edge("tools", "agent_node")
# After summarization: conversation turn is done
graph_builder.add_edge("summarize_node", END)
# ── Step 5: Compile ─────────────────────────────────────────
# Using MemorySaver for development.
# For production, swap this one line to SqliteSaver or PostgresSaver.
memory = MemorySaver()
shopbot = graph_builder.compile(checkpointer=memory)
In Colab or Jupyter, you can render the graph as a diagram:
from IPython.display import display, Image
from langchain_core.runnables.graph import MermaidDrawMethod
display(Image(
shopbot.get_graph().draw_mermaid_png(
draw_method=MermaidDrawMethod.API
)
))
Seeing a real flowchart of what you built is one of LangGraph's best features for learners. If the diagram looks wrong, the edges are wrong.
Module 7: entrypoint
This is where the agent actually runs. In a script it's the if __name__ == "__main__" block, and in an API server it's the route handler, though the pattern is the same either way.
# ============================================================
# MODULE 7: ENTRYPOINT
# ============================================================
def run_shopbot():
"""
Interactive command-line session with ShopBot.
Demonstrates: multi-turn conversation, tool use, and human-in-the-loop.
"""
print("=" * 55)
print(" ShopBot - Customer Support Agent")
print(" Powered by LangGraph")
print("=" * 55)
print("Type your message below. Type 'exit' to quit.")
print("Type 'state' to inspect what ShopBot currently remembers.\n")
# One config per session.
# thread_id is the session key - same ID = same memory thread.
# Change the ID to start a completely fresh conversation.
config = {"configurable": {"thread_id": "customer-session-001"}}
while True:
user_input = input("You: ").strip()
if not user_input:
continue
if user_input.lower() == "exit":
print("ShopBot: Thank you for contacting support. Have a great day!")
break
# ── Debug: inspect current state ────────────────────
if user_input.lower() == "state":
snapshot = shopbot.get_state(config)
print("\n[DEBUG] Current State:")
print(f" Messages in state : {len(snapshot.values.get('messages', []))}")
print(f" Customer name : {snapshot.values.get('customer_name', '(not set)')}")
print(f" Ticket category : {snapshot.values.get('ticket_category', '(not set)')}")
print(f" Summary : {snapshot.values.get('summary', '(none yet)')}")
print(f" Next node(s) : {snapshot.next}\n")
continue
# ── Normal message: invoke the graph ─────────────────
result = shopbot.invoke(
{"messages": [HumanMessage(content=user_input)]},
config=config,
)
# ── Check if graph paused for human approval ─────────
# This is how you detect that interrupt() was called inside review_refund.
while "__interrupt__" in result:
interrupt_data = result["__interrupt__"][0].value
print("\n" + "=" * 55)
print(" HUMAN APPROVAL REQUIRED")
print("=" * 55)
print(f" Customer : {interrupt_data.get('customer_name', 'Unknown')}")
print(f" Action : {interrupt_data.get('tool_being_called', 'refund')}")
print(f" Arguments : {interrupt_data.get('arguments', {})}")
print(f" Context : {interrupt_data.get('conversation_summary', 'N/A')}")
print("=" * 55)
print("Options: [a] Approve [r] Reject [e] Escalate")
human_choice = input("Your decision: ").strip().lower()
if human_choice == "a":
resume_value = "approve"
elif human_choice == "r":
resume_value = "reject"
elif human_choice == "e":
resume_value = "escalate"
else:
print("Invalid choice. Defaulting to reject.")
resume_value = "reject"
# Resume the graph with the human's decision.
# Command(resume=...) answers the pending interrupt() call.
result = shopbot.invoke(
Command(resume=resume_value),
config=config,
)
# ── Print the agent's final response ─────────────────
last_message = result["messages"][-1]
print(f"\nShopBot: {last_message.content}\n")
if __name__ == "__main__":
run_shopbot()
A real session
A sample conversation shows what to expect when you run it:
=======================================================
ShopBot — Customer Support Agent
Powered by LangGraph
=======================================================
Type your message below. Type 'exit' to quit.
Type 'state' to inspect what ShopBot currently remembers.
You: Hi, my name is Alex and I need help with order ORD-001
ShopBot: Hi Alex! I'd be happy to help you with order ORD-001.
I can see your order:
- Product: Wireless Headphones
- Status: Delivered on 2025-06-10
- Amount: $89.99
What can I help you with regarding this order?
You: I want to return it and get a refund
ShopBot: I've checked your order ORD-001 and it's eligible for a refund
since it's been delivered. Let me process that for you right away.
=======================================================
HUMAN APPROVAL REQUIRED
=======================================================
Customer : Alex
Action : process_refund
Arguments : {'order_id': 'ORD-001', 'reason': 'Customer requested return'}
Context : No summary yet
=======================================================
Options: [a] Approve [r] Reject [e] Escalate
Your decision: a
ShopBot: Great news, Alex! Your refund has been successfully processed.
Refund Reference: REF-ORD-001-143022
Amount: $89.99 will be returned to your original payment method within
3–5 business days.
Is there anything else I can help you with?
You: state
[DEBUG] Current State:
Messages in state : 6
Customer name : (not set)
Ticket category : refund_request
Summary : (none yet)
Next node(s) : ()
You: exit
ShopBot: Thank you for contacting support. Have a great day!
What the project demonstrates
The seven-module layout shows the architecture end to end: StateGraph, add_node, add_conditional_edges and compile, all in the right order. For memory, there's the summary field, the summarize_node function and the rolling summary pattern in which new messages are compressed and old ones deleted, plus a MemorySaver checkpointer that keeps history alive across turns. For human-in-the-loop, the review_refund node pauses with interrupt(), the entrypoint resumes with Command(resume=...), and a while "__interrupt__" in result loop copes with several approvals in a single turn, offering approve, reject or escalate.
Several things come from tooling. Real tools are declared with @tool, the prebuilt ToolNode runs them, and bind_tools attaches them to the model. The ReAct loop goes tools, agent node, tools and so on. And routing treats tools differently: safe ones go straight to tool_node, while sensitive ones pass through review_refund first.
Where to take it next
The project is deliberately complete yet minimal. Three extensions, easiest first:
- Better category detection. Right now
ticket_categorycomes from keyword scanning inagent_node. A cleaner route is asking the model to categorize the ticket explicitly with structured output, so look intowith_structured_output(). - SQLite persistence. Swap
MemorySaver()forSqliteSaver, a one-line change in Module 6, and add a startup prompt asking whether this is the same session or a new one. Conversations then survive restarts. - A complaint escalation path. Add a fourth node,
escalate_node, that fires whenticket_category == "complaint"and the customer's emotion is"FRUSTRATED". It callsinterrupt()to alert a human agent and hangs offroute_after_agentas a new branch.
Quick reference
New in this project are a handful of patterns. ToolNode(tools) is a pre-built node that executes the tool calls found in the model's last message. llm.bind_tools(tools) connects the tool list to the model, after which it can request a tool by returning a tool_calls field instead of plain text. @tool turns a Python function into something the model can call, with the docstring as what it reads to decide when to use it. The check hasattr(message, "tool_calls") and message.tool_calls is the standard way to see whether the last message contains a tool call, and hasattr should come first before touching the attribute. The while "__interrupt__" in result loop handles one or more pending interrupts, since each Command(resume=...) answers one and may trigger the next. Finally, shopbot.get_state(config) shows the full current state at any moment: .values gives the state dict and .next says what would run next, which is invaluable for debugging.
The gap between understanding a concept and writing code for it is real, and a small project that fits in your head yet exercises every tool in the box is how it closes. Here you have a checkpointer, real tools, an approval gate and a summarization trigger working together in one agent. Build it, break it on purpose, fix it, then extend it.