From 737e684ae8266107e0e3b20d1d23881891522d8c Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=D0=A0=D0=B8=D0=BD=D0=B0=D1=80=20=D0=9C=D0=B8=D1=80=D0=B7?=
 =?UTF-8?q?=D0=B0=D0=B3=D0=B8=D1=82=D0=BE=D0=B2?= <mirzagitov03@mail.ru>
Date: Thu, 11 Jun 2026 09:22:23 +0000
Subject: [PATCH] =?UTF-8?q?=D0=9F=D1=83=D0=B1=D0=BB=D0=B8=D0=BA=D0=B0?=
 =?UTF-8?q?=D1=86=D0=B8=D1=8F=20=D1=80=D0=B5=D1=88=D0=B5=D0=BD=D0=B8=D1=8F?=
 =?UTF-8?q?:=20add=20main.py?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 main.py | 91 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 91 insertions(+)
 create mode 100644 main.py

diff --git a/main.py b/main.py
new file mode 100644
index 0000000..db2f5d4
--- /dev/null
+++ b/main.py
@@ -0,0 +1,91 @@
+"""
+LangGraph planning agent example.
+"""
+
+from typing import TypedDict, List
+import os
+
+# State definition
+class PlanningState(TypedDict):
+    task: str
+    plan: List[str] | None
+    current_step: int
+    results: List[str]
+
+# LLM setup (OpenAI)
+from langchain_openai import ChatOpenAI
+llm = ChatOpenAI(model="gpt-4o-mini", temperature=0.2, api_key=os.getenv("OPENAI_API_KEY"))
+
+# Planning node
+async def planning(state: PlanningState) -> PlanningState:
+    prompt = (
+        f"Разбей задачу '{state['task']}' на 3–6 конкретных шагов.\n"
+        "Ответ в виде нумерованного списка, без лишних слов."
+    )
+    response = await llm.agenerate([prompt])
+    text = response.generations[0][0].text.strip()
+    # Parse numbered list
+    steps: List[str] = []
+    for line in text.splitlines():
+        if line.lstrip().startswith("1") or line.lstrip()[0].isdigit():
+            step = line.split('.', 1)[-1].strip()
+            if step:
+                steps.append(step)
+    return {
+        "task": state["task"],
+        "plan": steps,
+        "current_step": 0,
+        "results": [],
+    }
+
+# Execution node
+async def execution(state: PlanningState) -> PlanningState:
+    idx = state["current_step"]
+    step_text = state["plan"][idx]
+    prompt = f"Выполни шаг {idx+1}: {step_text}.\nОтвет в виде короткого абзаца." 
+    response = await llm.agenerate([prompt])
+    result = response.generations[0][0].text.strip()
+    new_results = state["results"] + [result]
+    return {
+        "task": state["task"],
+        "plan": state["plan"],
+        "current_step": idx + 1,
+        "results": new_results,
+    }
+
+# Condition node
+from langgraph import StateGraph, END
+
+def should_continue(state: PlanningState):
+    if state["current_step"] >= len(state["plan"]):
+        return "finish"
+    return "execute"
+
+# Graph definition
+workflow = StateGraph(PlanningState)
+workflow.add_node("planning", planning)
+workflow.add_node("execution", execution)
+workflow.set_entry_point("planning")
+workflow.add_conditional_edges(
+    "execution",
+    should_continue,
+    {
+        "execute": "execution",
+        "finish": END,
+    },
+)
+
+# Run example
+if __name__ == "__main__":
+    task = "Сравни Python и JavaScript"
+    initial_state: PlanningState = {"task": task, "plan": None, "current_step": 0, "results": []}
+    result = workflow.invoke(initial_state)
+    print("План:")
+    for i, step in enumerate(result["plan"]):
+        print(f"{i+1}. {step}")
+    print("\nШаги: ")
+    for i, res in enumerate(result["results"]):
+        print(f"[Шаг {i+1}] {res}\n")
+    print("Итог:")
+    final = llm.invoke("\n\nСводка всех результатов: " + "\n".join(result["results"]))
+    print(final)