fix: apply ruff formatter

.pre-commit-config.yaml

@@ -1,3 +1,4 @@
+repos:
 - repo: https://github.com/astral-sh/ruff-pre-commit
   # Ruff version.
   rev: v0.4.4

examples/frontend/app.py

@@ -2,9 +2,9 @@
 import sys
 
 import streamlit as st
+from utils import get_images
 
 from codeinterpreterapi import File
-from codeinterpreterapi.config import settings
 
 # Page configuration
 st.set_page_config(layout="wide")

server.py

@@ -2,15 +2,8 @@
     ConnectorImplCodeinterpreterApi,
 )
 from gui_agent_loop_core.gui_agent_loop_core import GuiAgentLoopCore
-from gui_agent_loop_core.schema.schema import (
-    GuiAgentInterpreterChatMessage,
-    GuiAgentInterpreterChatMessages,
-    GuiAgentInterpreterChatResponse,
-    GuiAgentInterpreterChatResponseAny,
-)
-from langchain_core.messages import SystemMessage
 
-from codeinterpreterapi import CodeInterpreterSession, settings
+from codeinterpreterapi import CodeInterpreterSession
 
 
 class CodeInterpreter(ConnectorImplCodeinterpreterApi):

src/codeinterpreterapi/agents/plan_and_execute/agent_executor.py

@@ -1,7 +1,7 @@
 from typing import List, Optional
 
 from langchain.agents.agent import AgentExecutor, AgentOutputParser
-from langchain.agents.structured_chat.base import StructuredChatAgent, create_structured_chat_agent
+from langchain.agents.structured_chat.base import create_structured_chat_agent
 from langchain.tools import BaseTool
 from langchain_core.callbacks import BaseCallbackManager
 from langchain_core.language_models import BaseLanguageModel

src/codeinterpreterapi/agents/plan_and_execute/prompts.py

@@ -1,7 +1,4 @@
-from textwrap import dedent
-
-from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder, PromptTemplate
-from langchain_experimental.tot.prompts import JSONListOutputParser
+from langchain_core.prompts import ChatPromptTemplate, MessagesPlaceholder
 
 SYSTEM_MESSAGE_TEMPLATE = '''Respond to the human as helpfully and accurately as possible. You have access to the following tools:
 

src/codeinterpreterapi/planners/planners.py

@@ -1,9 +1,7 @@
 from langchain import hub
-from langchain.agents import AgentExecutor, create_react_agent
+from langchain.agents import create_react_agent
 from langchain.base_language import BaseLanguageModel
-from langchain_core.runnables import Runnable, RunnablePassthrough
-from langchain_experimental.plan_and_execute import load_chat_planner
-from langchain_experimental.plan_and_execute.planners.base import LLMPlanner
+from langchain_core.runnables import Runnable
 
 SYSTEM_PROMPT_PLANNER = """
 Let's first understand the problem and devise a plan to solve the problem.

src/codeinterpreterapi/supervisors/supervisors.py

@@ -1,7 +1,10 @@
-from langchain import hub
+import getpass
+import os
+import platform
+
 from langchain.agents import AgentExecutor
 from langchain.chains.base import Chain
-from langchain_core.runnables import RunnablePassthrough
+from langchain_core.runnables import RunnableAssign, RunnablePassthrough
 from langchain_experimental.plan_and_execute.agent_executor import PlanAndExecute
 from langchain_experimental.plan_and_execute.planners.base import LLMPlanner
 
@@ -13,11 +16,21 @@ class MySupervisorChain(Chain):
 class CodeInterpreterSupervisor:
     @staticmethod
     def choose_supervisor(planner: LLMPlanner, executor: AgentExecutor, verbose: bool = False) -> MySupervisorChain:
+        # prompt
+        username = getpass.getuser()
+        current_working_directory = os.getcwd()
+        operating_system = platform.system()
+        info = f"[User Info]\nName: {username}\nCWD: {current_working_directory}\nOS: {operating_system}"
+        print("choose_supervisor info=", info)
+
         supervisor = PlanAndExecute(planner=planner, executor=executor, verbose=verbose)
-        prompt = hub.pull("nobu/chat_planner")
-        agent = create_react_agent(llm, [], prompt)
-        return agent
+        # prompt = hub.pull("nobu/chat_planner")
+        # agent = create_react_agent(llm, [], prompt)
+        # return agent
         # prompt = hub.pull("nobu/code_writer:0c56967d")
 
         supervisor_chain = RunnablePassthrough() | supervisor
         return supervisor_chain
+
+        supervisor_chain = RunnableAssign() | supervisor
+        return supervisor_chain

src/codeinterpreterapi/thoughts/thoughts.py

@@ -1,13 +1,9 @@
 from gui_agent_loop_core.thoughts.prompts import get_propose_prompt, get_propose_prompt_ja
-from langchain.agents import AgentExecutor
-from langchain.chains.base import Chain
 from langchain.prompts.base import BasePromptTemplate
-from langchain_experimental.plan_and_execute.agent_executor import PlanAndExecute
-from langchain_experimental.plan_and_execute.planners.base import LLMPlanner
 from langchain_experimental.pydantic_v1 import Field
 
 # from langchain_experimental.tot.prompts import get_cot_prompt, get_propose_prompt
-from langchain_experimental.tot.thought_generation import BaseThoughtGenerationStrategy, ProposePromptStrategy
+from langchain_experimental.tot.thought_generation import ProposePromptStrategy
 
 
 class MyProposePromptStrategy(ProposePromptStrategy):

src/codeinterpreterapi/tools/tools.py

@@ -1,8 +1,7 @@
 from langchain_community.tools.shell.tool import BaseTool, ShellTool
 from langchain_community.tools.tavily_search import TavilySearchResults
-from langchain_core.language_models import LLM, BaseLanguageModel, BaseLLM
+from langchain_core.language_models import BaseLanguageModel
 from langchain_core.tools import StructuredTool, Tool
-from langchain_experimental.chat_models.llm_wrapper import ChatWrapper
 from langchain_experimental.llm_bash.base import LLMBashChain
 
 from codeinterpreterapi.config import settings
@@ -65,7 +64,6 @@ def get_shell_v2(additional_tools: list[BaseTool], llm: BaseLanguageModel) -> li
              commands
              field required (type=value_error.missing)
         """
-        llm_runnable = ChatWrapper(llm=llm)
         bash_chain = LLMBashChain.from_llm(llm=llm)
         bash_tool = Tool(
             name="Bash", func=bash_chain.invoke, description="Executes bash commands in a terminal environment."