Using Tools with LangChain


import { ChatAnthropic } from "@langchain/anthropic";
import { ChatPromptTemplate } from "@langchain/core/prompts";
import { createToolCallingAgent } from "langchain/agents";
import { AgentExecutor } from "langchain/agents";
import { LangChainAdapter } from "@reacter/openapitools";
 
// Initialize the language model (Claude or OpenAI)
const llm = new ChatAnthropic({
  model: "claude-3-7-sonnet-20250219",
  apiKey: "your-anthropic-api-key",
  verbose: true,
});
 
// Initialize the adapter
const toolsAdapter = new LangChainAdapter(
  "your OpenAPI Tools Apikey - https://openapitools.com/dashboard/settings",
  {
    autoRefreshCount: 50,
  }
);
 
async function main() {
  // Define a prompt template with placeholders
  const prompt = ChatPromptTemplate.fromMessages([
    ["system", "You are a helpful assistant"],
    ["placeholder", "{chat_history}"],
    ["human", "{input}"],
    ["placeholder", "{agent_scratchpad}"],
  ]);
 
  // Get tools in LangChain format
  const tools = await toolsAdapter.getLangChainTools();
 
  // Create an agent with the LLM, tools, and prompt
  const agent = createToolCallingAgent({
    llm,
    tools,
    prompt,
  });
 
  // Create an executor to use the agent
  const agentExecutor = new AgentExecutor({
    agent,
    tools,
  });
 
  // Invoke the agent with a user query
  const res = await agentExecutor.invoke({
    input: "Can you generate an OTP for 9347994869?",
  });
 
  // Output the result
  console.log("Agent response:");
  console.log(res);
}
 
main().catch(console.error);

The LangChain integration provides a different approach to using tools by leveraging LangChain’s agent framework. Instead of manually handling the tool calls, the AgentExecutor takes care of the back-and-forth with the model. This is particularly useful when building applications where you want to simplify the interaction flow.

Key benefits of using the LangChain integration:

Simplified coding: The agent manages the conversation flow and tool execution
Standardized interface: Works with different LLM providers through LangChain’s interfaces
Access to LangChain features: Leverage LangChain’s memory, prompt templates, and other tools

To use different models, you can swap out the LLM implementation:


// Using OpenAI
import { ChatOpenAI } from "@langchain/openai";
 
const llm = new ChatOpenAI({
  model: "gpt-4o",
  apiKey: "your-openai-api-key",
  verbose: true,
});
 
// Rest of the code remains the same