OpenAI Agents SDK observability installation

Let AI instrument your LLM calls for you

Skip the manual setup — run this in your project and the wizard installs the SDK and wires up AI Observability for you.

Learn more
PostHog Wizard hedgehog

Contents

  1. Install the PostHog SDK

    Required

    Setting up analytics starts with installing the PostHog Python SDK.

    pip install posthog

  2. Install the OpenAI Agents SDK

    Required

    Install the OpenAI Agents SDK. PostHog instruments your agent runs by registering a tracing processor. The PostHog SDK does not proxy your calls.

    pip install openai-agents
    Proxy note

    These SDKs do not proxy your calls. They only fire off an async call to PostHog in the background to send the data. You can also use AI observability with other SDKs or our API, but you will need to capture the data in the right format. See the schema in the manual capture section for more details.

  3. Initialize PostHog tracing

    Required

    Initialize PostHog with your project token and host from your project settings. Then call instrument() to register PostHog tracing with the OpenAI Agents SDK. This automatically captures all agent traces, spans, and LLM generations.

    from posthog import Posthog
    from posthog.ai.openai_agents import instrument
    

    posthog = Posthog(
        "<ph_project_token>",
        host="https://us.i.posthog.com"
    )
    

    instrument(
        client=posthog,
        distinct_id=lambda trace: (trace.metadata or {}).get("posthog_distinct_id"),
        privacy_mode=False, # optional
        groups={"company": "company_id_in_your_db"}, # optional
    )

    Note: If you want to capture LLM events anonymously, do not pass a distinct ID — here or per run. See our docs on anonymous vs identified events to learn more.

  4. Run your agents

    Required

    Run your OpenAI agents as normal. PostHog automatically captures $ai_generation events for LLM calls and $ai_span events for agent execution, tool calls, and handoffs. Pass the user and conversation on the run's RunConfig:

    • group_id groups the run's traces into a conversation — it becomes $ai_session_id.
    • trace_metadata["posthog_distinct_id"] attributes the run's events to a user — the distinct_id lambda from the previous step reads it off each trace. Any other trace_metadata keys land on the trace as $ai_trace_metadata.

    The example below defines a tool and lets the agent call it.

    from agents import Agent, Runner, RunConfig, function_tool
    

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

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

    result = Runner.run_sync(
        agent,
        "What's the weather in Paris?",
        run_config=RunConfig(
            group_id="conversation_abc",
            trace_metadata={"posthog_distinct_id": "user_123"},
        ),
    )
    print(result.final_output)

    You can expect captured $ai_generation events to have the following properties:

    PropertyDescription
    $ai_modelThe specific model, like gpt-5-mini or claude-4-sonnet
    $ai_latencyThe latency of the LLM call in seconds
    $ai_time_to_first_tokenTime to first token in seconds (streaming only)
    $ai_toolsTools and functions available to the LLM
    $ai_inputList of messages sent to the LLM
    $ai_input_tokensThe number of tokens in the input (often found in response.usage)
    $ai_output_choicesList of response choices from the LLM
    $ai_output_tokensThe number of tokens in the output (often found in response.usage)
    $ai_total_cost_usdThe total cost in USD (input + output)
    [...]See full list of properties

  5. Multi-agent and tool usage

    Optional

    PostHog captures the full trace hierarchy for complex agent workflows, including handoffs between multiple agents.

    from agents import Agent, Runner, function_tool
    

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

    weather_agent = Agent(
        name="WeatherAgent",
        instructions="You help with weather queries.",
        tools=[get_weather]
    )
    

    triage_agent = Agent(
        name="TriageAgent",
        instructions="Route weather questions to the weather agent.",
        handoffs=[weather_agent]
    )
    

    result = Runner.run_sync(triage_agent, "What's the weather in San Francisco?")

    This captures:

    • Agent spans for TriageAgent and WeatherAgent
    • Handoff spans showing the routing between agents
    • Tool spans for get_weather function calls
    • Generation spans for all LLM calls

    As with the single-agent example above, PostHog captures every span in that list automatically. You write no extra code for the handoff itself.

  6. Verify traces and generations

    Recommended
    Confirm LLM events are being sent to PostHog

    Let's make sure LLM events are being captured and sent to PostHog. Under AI Observability, you should see rows of data appear in the Traces and Generations tabs.


    LLM generations in PostHog
    Check for LLM events in PostHog

  7. Next steps

    Recommended

    Now that you're capturing AI conversations, continue with the resources below to learn what else AI Observability enables within the PostHog platform.

    ResourceDescription
    BasicsLearn the basics of how LLM calls become events in PostHog.
    GenerationsRead about the $ai_generation event and its properties.
    TracesExplore the trace hierarchy and how to use it to debug LLM calls.
    SpansReview spans and their role in representing individual operations.
    Anaylze LLM performanceLearn how to create dashboards to analyze LLM performance.

Community questions

Was this page useful?