LangChain

Trace LangChain chain, LLM, and tool invocations by attaching an AxonPushCallbackHandler to your chain’s callback list. Every lifecycle event ships to AxonPush in real time.

Installation

npm install @axonpush/sdk @langchain/core

Setup

import { AxonPush } from "@axonpush/sdk";
import { AxonPushCallbackHandler } from "@axonpush/sdk/integrations/langchain";

const client = new AxonPush({
  apiKey: process.env.AXONPUSH_API_KEY!,
  tenantId: process.env.AXONPUSH_TENANT_ID!,
});

const handler = new AxonPushCallbackHandler({
  client,
  channelId: 1,
  agentId: "my-agent",
});

Usage

import { ChatOpenAI } from "@langchain/openai";
import { RunnableSequence } from "@langchain/core/runnables";

const chain = RunnableSequence.from([prompt, new ChatOpenAI()]);
const result = await chain.invoke(input, { callbacks: [handler] });

Events Traced

Event	When
chain.start	A chain or runnable begins execution
chain.end	A chain or runnable completes successfully
chain.error	A chain or runnable throws an error
llm.start	An LLM call begins
llm.end	An LLM call completes
llm.token	A streaming token is received
llm.error	An LLM call fails
tool.*.start	A tool invocation begins
tool.end	A tool invocation completes
tool.error	A tool invocation fails