I bridge the gap between Customer Psychology, Applied AI, and Production Code.

from typing import TypedDict, Annotated, Sequence
import operator
from langgraph.graph import StateGraph, END
from langchain_core.messages import BaseMessage, HumanMessage, SystemMessage
from langchain_google_genai import ChatGoogleGenerativeAI

class AgentState(TypedDict):
    product_sku: str
    target_url: str
    scraped_price: float
    current_inventory_margin: float
    reprice_action_needed: bool
    final_output: str

llm = ChatGoogleGenerativeAI(model="gemini-2.5-flash", temperature=0.1)

async def scrape_competitor(state: AgentState):
    # Simulated scraper node
    print(f"Scraping {state['target_url']}...")
    return {"scraped_price": 79.99}

async def evaluate_margins(state: AgentState):
    prompt = f"""Compare scraped price {state['scraped_price']} with margin {state['current_inventory_margin']}.
    Should we match price? Output YES or NO with justification."""
    response = await llm.ainvoke([HumanMessage(content=prompt)])
    reprice = "YES" in response.content.upper()
    return {"reprice_action_needed": reprice, "final_output": response.content}

async def execute_action(state: AgentState):
    if state["reprice_action_needed"]:
        print("Emitting webhook to Shopify & Slack: Price updated!")
    return {"final_output": "Workflow completed successfully."}

workflow = StateGraph(AgentState)
workflow.add_node("scraper", scrape_competitor)
workflow.add_node("evaluator", evaluate_margins)
workflow.add_node("executor", execute_action)

workflow.set_entry_point("scraper")
workflow.add_edge("scraper", "evaluator")
workflow.add_edge("evaluator", "executor")
workflow.add_edge("executor", END)

app = workflow.compile()