Skip to main content
  1. Sign up — Create an account at platform.respan.ai
  2. Create an API key — Generate one on the API keys page
  3. Add credits or a provider key — Add credits on the Credits page or connect your own provider key on the Integrations page

​
What is OpenAI Agents?

This integration is for agent tracing. For the OpenAI gateway integration, see OpenAI SDK.
The OpenAI Agents SDK is a lightweight framework for building multi-agent workflows with tools, handoffs, and guardrails. Respan captures every agent run as a trace so you can debug and optimize your workflows.
Agent tracing visualization

​
Setup

1

Install packages

pip install openai-agents respan-exporter-openai-agents
2

Set environment variables

OPENAI_API_KEY=your-openai-api-key
RESPAN_API_KEY=your-respan-api-key
RESPAN_OAIA_TRACING_ENDPOINT=https://api.respan.ai/api/openai/v1/traces/ingest
3

Initialize the exporter and run an agent

import asyncio
import os
from agents import Agent, Runner
from agents.tracing import set_trace_processors, trace
from respan_exporter_openai_agents import RespanTraceProcessor

set_trace_processors([
    RespanTraceProcessor(
        api_key=os.getenv("RESPAN_API_KEY"),
        endpoint=os.getenv("RESPAN_OAIA_TRACING_ENDPOINT"),
    ),
])

async def main():
    agent = Agent(
        name="Assistant",
        instructions="You only respond in haikus.",
    )

    with trace("Hello world test"):
        result = await Runner.run(agent, "Tell me about recursion in programming.")
        print(result.final_output)

asyncio.run(main())
4

View your trace

Open the Traces page in the Respan dashboard. Your trace should appear within a few seconds.

​
Configuration

The Python exporter accepts these constructor parameters:
ParameterRequiredDescription
api_keyYesYour Respan API key. Falls back to RESPAN_API_KEY.
endpointYesIngest endpoint URL. Falls back to RESPAN_OAIA_TRACING_ENDPOINT.
The JavaScript exporter reads RESPAN_API_KEY and RESPAN_BASE_URL from environment variables. You can also pass metadata on a per-trace basis: 
await withTrace("My Trace", async () => {
  return run(agent, "Hello");
}, {
  metadata: { foo: "bar" }
});

​
Tool calls

Agents can use tools to perform specific tasks. Tool calls are automatically captured in the trace. 
import asyncio
import os
from agents import Agent, Runner, function_tool
from agents.tracing import set_trace_processors
from respan_exporter_openai_agents import RespanTraceProcessor

set_trace_processors([
    RespanTraceProcessor(
        api_key=os.getenv("RESPAN_API_KEY"),
        endpoint=os.getenv("RESPAN_OAIA_TRACING_ENDPOINT"),
    ),
])

@function_tool
def get_weather(city: str) -> str:
    return f"The weather in {city} is sunny."

agent = Agent(
    name="Hello world",
    instructions="You are a helpful agent.",
    tools=[get_weather],
)

async def main():
    result = await Runner.run(agent, input="What's the weather in Tokyo?")
    print(result.final_output)

asyncio.run(main())

​
Handoffs

Agents can hand off conversations to other specialized agents. Handoff events are captured in the trace. 
import asyncio
import os
from agents import Agent, Runner
from agents.tracing import set_trace_processors
from respan_exporter_openai_agents import RespanTraceProcessor

set_trace_processors([
    RespanTraceProcessor(
        api_key=os.getenv("RESPAN_API_KEY"),
        endpoint=os.getenv("RESPAN_OAIA_TRACING_ENDPOINT"),
    ),
])

spanish_agent = Agent(
    name="Spanish agent",
    instructions="You only speak Spanish.",
)

english_agent = Agent(
    name="English agent",
    instructions="You only speak English",
)

triage_agent = Agent(
    name="Triage agent",
    instructions="Handoff to the appropriate agent based on the language of the request.",
    handoffs=[spanish_agent, english_agent],
)

async def main():
    result = await Runner.run(triage_agent, input="Hola, ¿cómo estás?")
    print(result.final_output)

asyncio.run(main())

​
Observability

With this integration, Respan auto-captures:
  • Agent runs — each agent execution as a span
  • LLM calls — model, input/output messages, token usage
  • Tool calls — function name, arguments, return value
  • Handoffs — which agent transferred control and to whom
  • Errors — failed runs and error details
View traces on the Traces page.
OpenAI Agents SDK trace in Respan