Agent工作流对比：LangChain vs AutoGen vs n8n

from langchain.memory import ConversationBufferMemory
from langchain.chat_models import ChatOpenAI
from langchain.chains import ConversationChain

# 创建记忆组件
memory = ConversationBufferMemory(
    memory_key="chat_history",
    return_messages=True
)

# 创建对话链
llm = ChatOpenAI(temperature=0)
conversation = ConversationChain(
    llm=llm,
    memory=memory,
    verbose=True
)

# 使用记忆进行对话
response = conversation.predict(input="你好，我是小明")
print(response)

response = conversation.predict(input="你还记得我的名字吗？")
print(response)

from langchain.vectorstores import Chroma
from langchain.embeddings import OpenAIEmbeddings
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain.document_loaders import TextLoader

# 加载文档
loader = TextLoader("data.txt")
documents = loader.load()

# 分割文档
text_splitter = RecursiveCharacterTextSplitter(
    chunk_size=1000,
    chunk_overlap=200
)
texts = text_splitter.split_documents(documents)

# 创建向量数据库
embeddings = OpenAIEmbeddings()
vectorstore = Chroma.from_documents(
    documents=texts,
    embedding=embeddings
)

# 检索相关文档
query = "什么是机器学习？"
docs = vectorstore.similarity_search(query)
print(docs[0].page_content)

import autogen
from autogen import AssistantAgent, UserProxyAgent

# 配置LLM
config_list = [
    {
        "model": "gpt-4",
        "api_key": "your-api-key"
    }
]

# 创建助手Agent
assistant = AssistantAgent(
    name="assistant",
    llm_config={"config_list": config_list},
    system_message="你是一个有用的AI助手。"
)

# 创建用户代理
user_proxy = UserProxyAgent(
    name="user_proxy",
    human_input_mode="NEVER",
    max_consecutive_auto_reply=10,
    llm_config={"config_list": config_list}
)

# 开始对话
user_proxy.initiate_chat(
    assistant,
    message="请帮我分析一下当前AI技术的发展趋势。"
)

import autogen
from autogen import AssistantAgent, UserProxyAgent

# 创建带记忆的Agent
assistant = AssistantAgent(
    name="assistant",
    llm_config={"config_list": config_list},
    system_message="你是一个有记忆的AI助手。",
    memory=True  # 启用记忆功能
)

user_proxy = UserProxyAgent(
    name="user_proxy",
    human_input_mode="NEVER",
    max_consecutive_auto_reply=10,
    llm_config={"config_list": config_list},
    memory=True
)

# 多轮对话，Agent会记住之前的对话
user_proxy.initiate_chat(assistant, "我的名字是张三")
user_proxy.initiate_chat(assistant, "你还记得我的名字吗？")

// n8n工作流配置示例
{
  "name": "AI知识库工作流",
  "nodes": [
    {
      "parameters": {
        "url": "https://api.openai.com/v1/chat/completions",
        "authentication": "predefinedCredentialType",
        "nodeCredentialType": "openAiApi",
        "method": "POST",
        "sendHeaders": true,
        "headerParameters": {
          "parameters": [
            {
              "name": "Content-Type",
              "value": "application/json"
            }
          ]
        },
        "sendBody": true,
        "bodyParameters": {
          "parameters": [
            {
              "name": "model",
              "value": "gpt-4"
            },
            {
              "name": "messages",
              "value": "={{ $json.messages }}"
            }
          ]
        }
      },
      "id": "openai-node",
      "name": "OpenAI",
      "type": "n8n-nodes-base.openAi",
      "typeVersion": 1,
      "position": [240, 300]
    }
  ]
}

// 数据处理节点配置
{
  "parameters": {
    "jsCode": "// 处理知识库数据\nconst items = $input.all();\nconst processedData = [];\n\nfor (const item of items) {\n  const text = item.json.content;\n  const chunks = text.match(/.{1,1000}/g) || [];\n  \n  for (const chunk of chunks) {\n    processedData.push({\n      content: chunk,\n      metadata: {\n        source: item.json.source,\n        timestamp: new Date().toISOString()\n      }\n    });\n  }\n}\n\nreturn processedData.map(data => ({ json: data }));"
  },
  "id": "code-node",
  "name": "Process Data",
  "type": "n8n-nodes-base.code",
  "typeVersion": 2,
  "position": [460, 300]
}