Skip to main content

​
Integrating Agno with Traceloop

Traceloop provides an LLM observability platform built on OpenLLMetry, an open-source OpenTelemetry extension. By integrating Agno with Traceloop, you can automatically trace agent execution, team workflows, tool calls, and token usage metrics.

​
Prerequisites

  1. Install Dependencies Ensure you have the necessary packages installed: 
    uv pip install agno openai traceloop-sdk
  2. Setup Traceloop Account
  3. Set Environment Variables Configure your environment with the Traceloop API key: 
    export TRACELOOP_API_KEY=<your-api-key>

​
Sending Traces to Traceloop

  • ​
    Example: Basic Agent Instrumentation

Initialize Traceloop at the start of your application. The SDK automatically instruments Agno agent execution. 
from traceloop.sdk import Traceloop
from agno.agent import Agent
from agno.models.openai import OpenAIResponses

# Initialize Traceloop - must be called before creating agents
Traceloop.init(app_name="agno_agent")

# Create and configure the agent
agent = Agent(
    name="Assistant",
    model=OpenAIResponses(id="gpt-5.2"),
    description="A helpful assistant",
    instructions=["Be concise and helpful"],
)

# Agent execution is automatically traced
response = agent.run("What is the capital of France?")
print(response.content)

  • ​
    Example: Development Mode (Disable Batching)

For local development, disable batching to see traces immediately: 
from traceloop.sdk import Traceloop
from agno.agent import Agent
from agno.models.openai import OpenAIResponses

# Disable batching for immediate trace visibility during development
Traceloop.init(app_name="agno_dev", disable_batch=True)

# Create and configure the agent
agent = Agent(
    name="DevAgent",
    model=OpenAIResponses(id="gpt-5.2"),
)

agent.print_response("Hello, world!")

  • ​
    Example: Multi-Agent Team Tracing

Team execution is automatically traced, showing the coordination between multiple agents: 
from traceloop.sdk import Traceloop
from agno.agent import Agent
from agno.models.openai import OpenAIResponses
from agno.team import Team

Traceloop.init(app_name="agno_team")

researcher = Agent(
    name="Researcher",
    role="Research Specialist",
    model=OpenAIResponses(id="gpt-5.2"),
    instructions=["Research topics thoroughly and provide factual information"],
    debug_mode=True,
)

writer = Agent(
    name="Writer",
    role="Content Writer",
    model=OpenAIResponses(id="gpt-5.2"),
    instructions=["Write clear, engaging content based on research"],
    debug_mode=True,
)

team = Team(
    name="ContentTeam",
    members=[researcher, writer],
    model=OpenAIResponses(id="gpt-5.2"),
    debug_mode=True,
)

# Team execution creates parent span with child spans for each agent
result = team.run("Write a brief overview of OpenTelemetry observability")
print(result.content)

  • ​
    Example: Using Workflow Decorators

Use the @workflow decorator to create custom spans for organizing your traces: 
from traceloop.sdk import Traceloop
from traceloop.sdk.decorators import workflow
from agno.agent import Agent
from agno.models.openai import OpenAIResponses

Traceloop.init(app_name="agno_workflows")

agent = Agent(
    name="AnalysisAgent",
    model=OpenAIResponses(id="gpt-5.2"),
    debug_mode=True,
)

@workflow(name="data_analysis_pipeline")
def analyze_data(query: str) -> str:
    """Custom workflow that wraps agent execution."""
    response = agent.run(query)
    return response.content

# The workflow decorator creates a parent span
result = analyze_data("Analyze the benefits of observability in AI systems")
print(result)

  • ​
    Example: Async Agent with Tools

Async agent execution is fully supported with automatic tool call tracing: 
import asyncio
from traceloop.sdk import Traceloop
from agno.agent import Agent
from agno.models.openai import OpenAIResponses

Traceloop.init(app_name="agno_async")

def get_weather(city: str) -> str:
    """Get the weather for a city."""
    return f"The weather in {city} is sunny, 72°F"

agent = Agent(
    name="WeatherAgent",
    model=OpenAIResponses(id="gpt-5.2"),
    tools=[get_weather],
    debug_mode=True,
)

async def main():
    # Async execution is automatically traced
    response = await agent.arun("What's the weather in San Francisco?")
    print(response.content)

asyncio.run(main())

​
Notes

  • Initialization: Call Traceloop.init() before creating any agents to ensure proper instrumentation.
  • Development Mode: Use disable_batch=True during development for immediate trace visibility.
  • Async Support: Both sync (run()) and async (arun()) methods are fully instrumented.
  • Privacy Control: Set TRACELOOP_TRACE_CONTENT=false to disable logging of prompts and completions.