ati-integrations-langchain 0.1.0

ATI LangChain integration: emit agent-aware OpenTelemetry spans via LangChain callbacks

ATI Integration for LangChain

This package provides OpenTelemetry instrumentation for LangChain agents using IOcane ATI.

Installation

pip install ati-integrations-langchain

Usage

from ati_langchain import LangChainInstrumentor
from langchain.chat_models import ChatOpenAI
from langchain.agents import initialize_agent, AgentType

# 1. Enable Instrumentation
# This returns a callback handler that captures ATI spans.
handler = LangChainInstrumentor().instrument()

# 2. Use in LangChain
llm = ChatOpenAI(temperature=0, callbacks=[handler])
# or pass globally if using a CallbackManager

Configuration

Configure the instrumentation via environment variables:

Variable	Description	Default
ATI_CAPTURE_PAYLOADS	Capture input/output payloads (redacted)	false
ATI_DEBUG	Enable debug logging	false

Features

• Captures LLM calls (ati.span.type=llm)
• Captures Tool usage (ati.span.type=tool)
• Captures Chain/Agent execution steps