LangChain Integration

Add full observability and governance to your LangChain agents — zero code changes required.

Install

pip install nyraxis-sdk

Quick start

import nyraxis_sdk
from langchain_openai import ChatOpenAI

nyraxis_sdk.init(
    api_key="nyx_your_api_key",
    agent_name="my-langchain-agent",
)

# Use LangChain as normal — all calls are auto-traced
llm = ChatOpenAI(model="gpt-4o")
response = llm.invoke("Summarize the latest AI news")

nyraxis_sdk.shutdown()

What gets captured

• Every LLM call — model, prompt, completion, token counts, latency
• Tool calls — name, parameters, result
• LCEL chains and LangGraph agents — full span tree
• Cost auto-calculated from token counts + model pricing table
• Governance policies evaluated in real-time (PII, prompt injection, cost limits)

Trace grouping

from nyraxis_sdk import workflow, task

@workflow(name="research-agent")
def run_agent(query):
    return summarize(query)

@task(name="summarize")
def summarize(text):
    llm = ChatOpenAI(model="gpt-4o")
    return llm.invoke(f"Summarize: {text}")

View traces at Dashboard → Traces.

Other integrations

• CrewAI
• LlamaIndex
• Vercel AI SDK