fix: main.py — Создайть просто AI агент на Python с применением langchain

This commit is contained in:
2026-07-02 07:06:28 +00:00
parent 6489fb5fc0
commit 7da640e2c7
+53 -53
View File
@@ -1,17 +1,18 @@
import os
import asyncio
from typing import Any, Dict, List
from typing import List, Dict, Any
from pydantic import SecretStr
from langchain_openai import ChatOpenAI
from langchain_core.messages import HumanMessage, AIMessage, ToolMessage
from langchain_core.messages import HumanMessage, BaseMessage
from langchain.tools import tool
from deepagents import create_deep_agent
from deepagents.backends import CompositeBackend, LocalShellBackend, FilesystemBackend
from deepagents.backends import FilesystemBackend, LocalShellBackend, CompositeBackend
# ----------------------------------------------------------------------
# LLM configuration (OpenRouter)
# Configuration
# ----------------------------------------------------------------------
# LLM - OpenRouter (free tier). The API key must be stored in the environment.
llm = ChatOpenAI(
model="openai/gpt-oss-20b:free",
base_url="https://openrouter.ai/api/v1",
@@ -19,9 +20,8 @@ llm = ChatOpenAI(
temperature=0.7,
)
# ----------------------------------------------------------------------
# Backend for sub-agents (allows file operations and shell commands)
# ----------------------------------------------------------------------
# Backend for the agents - a simple composite that allows file operations
# and execution of shell commands inside a sandboxed workspace.
backend = CompositeBackend(
[
LocalShellBackend(workspace_dir="./workspace"),
@@ -30,90 +30,90 @@ backend = CompositeBackend(
)
# ----------------------------------------------------------------------
# Sub-agent that generates a realistic price table for a product
# Sub-agent: price generator
# ----------------------------------------------------------------------
def create_price_subagent() -> Any:
def _create_price_subagent() -> Any:
"""
Returns a deep agent that, given a product and a city, produces a markdown
table with product, price and store. The prompt forces the model to fabricate
plausible data based on typical market prices.
Creates a lightweight sub-agent that, given a product and a city,
returns a markdown table with a plausible price and a store name.
The sub-agent re-uses the same LLM and backend as the main agent.
"""
system_prompt = (
"You are a price-generation sub-agent. Given a product name and a city, "
"return a markdown table with columns: Продукт, Цена (руб.), Магазин. "
"Fabricate realistic prices based on typical Russian market data. "
"Do not add any extra commentary, only the table."
)
subagent = create_deep_agent(
model=llm,
tools=[], # no external tools needed for this simple sub-agent
tools=[], # No additional tools are required for price generation
backend=backend,
system_prompt=system_prompt,
system_prompt=(
"You are a price-estimation sub-agent. "
"Given a product name and a city, generate a realistic price "
"in Russian rubles and suggest a typical store. "
"Return the result as a markdown table with columns: "
"`Продукт`, `Цена (руб.)`, `Магазин`."
),
)
return subagent
price_subagent = create_price_subagent()
_price_subagent = _create_price_subagent()
# ----------------------------------------------------------------------
# Tool that calls the sub-agent
# ----------------------------------------------------------------------
@tool
def get_price(product: str, city: str) -> str:
"""
Generate a realistic price for the given product in the specified city.
Returns a markdown table with columns: Продукт, Цена (руб.), Магазин.
Estimate the price of a product in a given city.
The function creates a sub-agent that returns a markdown table:
| Продукт | Цена (руб.) | Магазин |
"""
# Build the prompt for the sub-agent
prompt = f"Продукт: {product}\nГород: {city}"
# Invoke the sub-agent synchronously (deepagents also supports async,
# but a simple sync call keeps the example straightforward)
prompt = HumanMessage(
content=f"Продукт: {product}\nГород: {city}\nСгенерируй цену."
)
# Invoke the sub-agent asynchronously and wait for the result
result = asyncio.run(
price_subagent.ainvoke(
{"messages": [HumanMessage(content=prompt)]},
_price_subagent.ainvoke(
{"messages": [prompt]},
{"configurable": {"thread_id": f"price-{product}-{city}"}},
)
)
# The sub-agent returns a list of messages; the last one contains the table
final_message = result["messages"][-1]
if isinstance(final_message, AIMessage):
return final_message.content
elif isinstance(final_message, ToolMessage):
return final_message.content
else:
return str(final_message)
return final_message.content if isinstance(final_message, BaseMessage) else str(final_message)
# ----------------------------------------------------------------------
# Main shopping-list agent
# Main agent: shopping list planner
# ----------------------------------------------------------------------
shopping_agent = create_deep_agent(
main_agent = create_deep_agent(
model=llm,
tools=[get_price],
backend=backend,
system_prompt="Ты помощник по планированию покупок.",
)
def format_message(msg: Any) -> str:
"""Human-readable representation of a message or tool call."""
if isinstance(msg, (HumanMessage, AIMessage)):
def format_message(msg: BaseMessage) -> str:
"""
Convert a LangChain message to a readable string.
Handles normal text messages and tool calls.
"""
if hasattr(msg, "content") and msg.content:
return msg.content
if isinstance(msg, ToolMessage):
return f"{msg.name}({msg.args}) -> {msg.content}"
# Fallback for generic dict-like messages
# Tool call representation
if hasattr(msg, "tool_calls") and msg.tool_calls:
call = msg.tool_calls[0]
return f"{call['name']}({call['args']})"
name = call["name"]
args = ", ".join(f"{k}={v!r}" for k, v in call["args"].items())
return f"{name}({args})"
return str(msg)
async def main() -> None:
user_query = "Помоги составить список покупок: молоко, хлеб, яблоки. Я нахожусь в Казани."
result = await shopping_agent.ainvoke(
user_query = (
"Помоги составить список покупок: молоко, хлеб, яблоки. "
"Я нахожусь в Казани."
)
result = await main_agent.ainvoke(
{"messages": [HumanMessage(content=user_query)]},
{"configurable": {"thread_id": "shopping-session-1"}},
)
# Print the whole chain of messages
for i, message in enumerate(result["messages"]):
print(f"--- Message {i + 1} ---")
print(format_message(message))
# Print the whole conversation chain
for i, msg in enumerate(result["messages"], start=1):
print(f"--- Message {i} ---")
print(format_message(msg))
print()
if __name__ == "__main__":