AutoGen

Trace AutoGen multi-agent conversations with OpenInference and send spans to Arize AX for LLM observability.

AutoGen is Microsoft’s framework for building multi-agent LLM applications — agents that converse with each other to solve tasks. Arize AX captures every AutoGen run — assistant turns, user-proxy auto-replies, and the underlying LLM calls — via the openinference-instrumentation-autogen package, paired with openinference-instrumentation-openai for full visibility into each LLM call.

AutoGen Tracing Tutorial (Google Colab)

This guide covers the original AutoGen API (autogen.AssistantAgent / UserProxyAgent / initiate_chat). For the newer AgentChat API (autogen-agentchat package), see the AutoGen AgentChat tracing guide.

Prerequisites

• Python 3.10+
• An Arize AX account (sign up)
• An OPENAI_API_KEY from the OpenAI Platform

Launch Arize AX

1. Sign in to your Arize AX account.
2. From Space Settings, copy your Space ID and API Key. You will set them as ARIZE_SPACE_ID and ARIZE_API_KEY below.

Install

pip install arize-otel \
  openinference-instrumentation-autogen \
  openinference-instrumentation-openai \
  autogen openai

Configure credentials

export ARIZE_SPACE_ID="<your-space-id>"
export ARIZE_API_KEY="<your-api-key>"
export ARIZE_PROJECT_NAME="autogen-tracing-example"
export OPENAI_API_KEY="<your-openai-api-key>"

Setup tracing

# instrumentation.py
import os

from arize.otel import register
from openinference.instrumentation.autogen import AutogenInstrumentor
from openinference.instrumentation.openai import OpenAIInstrumentor

tracer_provider = register(
    space_id=os.environ["ARIZE_SPACE_ID"],
    api_key=os.environ["ARIZE_API_KEY"],
    project_name=os.environ["ARIZE_PROJECT_NAME"],
    set_global_tracer_provider=True,
)

OpenAIInstrumentor().instrument(tracer_provider=tracer_provider)
AutogenInstrumentor().instrument()
print("Arize AX tracing initialized for AutoGen.")

Run AutoGen

# example.py

# Importing instrumentation first ensures tracing is set up
# before `autogen` is imported.
from instrumentation import tracer_provider

import autogen

# OpenAI client reads OPENAI_API_KEY from the environment.
config_list = [{"model": "gpt-5.5"}]

assistant = autogen.AssistantAgent(
    name="assistant",
    llm_config={"config_list": config_list},
)

user = autogen.UserProxyAgent(
    name="user",
    human_input_mode="NEVER",
    max_consecutive_auto_reply=1,
    code_execution_config=False,
    is_termination_msg=lambda m: "TERMINATE" in m.get("content", "").upper(),
)

result = user.initiate_chat(
    assistant,
    message=(
        "Why is the ocean salty? Answer in two sentences, "
        "then write TERMINATE on a new line."
    ),
)
print(result.summary)

Expected output

Arize AX tracing initialized for AutoGen.
The ocean is salty because rivers continuously dissolve mineral salts from rocks and soil and carry them to the sea, where they accumulate over millions of years. Water leaves the ocean through evaporation but the salts remain, steadily concentrating until reaching today's roughly 3.5% salinity.

Verify in Arize AX

1. Open your Arize AX space and select project autogen-tracing-example.
2. You should see a new trace within ~30 seconds containing an Autogen parent span wrapping OpenAI ChatCompletion LLM child spans with the prompt, response, and token usage attached.
3. If no traces appear, see Troubleshooting.

Check from the skill, CLI, or SDK

Confirm spans are actually reaching your Arize AX project. Use whichever fits your workflow — the skill and CLI work for any framework; the SDK check is shown for each language.

Arize skill (agent)

Install the Arize Skills plugin and let your coding agent check for you:

npx skills add Arize-ai/arize-skills

Then prompt your agent:

Use the arize-trace skill to export and analyze recent traces from my project. Confirm spans are arriving, and summarize any errors or latency issues.

AX CLI

Export recent spans for your project — any rows mean traces are landing:

ax spans export "$ARIZE_PROJECT_NAME" --space "$ARIZE_SPACE_ID" \
  --limit 5 --stdout | jq 'length'

A non-zero count confirms spans reached Arize AX. Run ax auth login first if you have not authenticated. See the ax spans reference.

SDK

Query the project’s spans and check that at least one came back.

Python

import os
from arize import ArizeClient

client = ArizeClient(api_key=os.environ["ARIZE_API_KEY"])
resp = client.spans.list(
    project=os.environ["ARIZE_PROJECT_NAME"],
    space=os.environ["ARIZE_SPACE_ID"],
    limit=5,
)
count = len(resp.spans)
print(
    f"{count} span(s) found" if count else "No spans yet — recheck setup"
)

SDK span references: Python · TypeScript · Go.

Troubleshooting

• No traces in Arize AX. Confirm ARIZE_SPACE_ID and ARIZE_API_KEY are set in the same shell that runs example.py. Enable OpenTelemetry debug logs with export OTEL_LOG_LEVEL=debug and re-run.
• AutoGen spans missing but OpenAI spans present. AutogenInstrumentor().instrument(...) must run before any autogen import. Make sure instrumentation.py is the first import in your entry point.
• 401 from OpenAI. Verify OPENAI_API_KEY is set and has access to gpt-5.5. Swap for a model your key can call.
• Conversation never terminates. AutoGen’s UserProxyAgent keeps replying until max_consecutive_auto_reply is reached or is_termination_msg returns True. Tighten the termination check or lower the cap.

Resources

AutoGen 0.2 Documentation

OpenInference AutoGen Instrumentor

AutoGen AgentChat Tracing (newer API)