# Openai agents sdk evals

The OpenAI Agents SDK is a small runtime for building agents. An agent is a model with
instructions and tools. A handoff passes the conversation to another agent. Guardrails run checks
on input and output. Sessions carry history between turns. The loop around all of this is thin,
which helps evaluation, because there's very little framework behaviour between what you wrote
and what ran.

Tracing is built in and on by default. One run is one trace, and the SDK already splits it into
the pieces you'd want to grade: a span for each agent, each model call, each tool call, each
handoff, and each guardrail. Two of those are worth more than they look. A handoff span records
the moment control moved from one agent to another and what went with it, which is where most
multi-agent failures start. A guardrail span records that a check ran and what it decided. That
is a separate fact from the model's output, and the rate at which guardrails trip is worth
tracking on its own.

Getting the spans out of OpenAI's backend takes a trace processor. There is no OpenTelemetry
exporter in the SDK. add_trace_processor sends spans to a second destination alongside OpenAI's,
set_trace_processors replaces it, and OPENAI_AGENTS_DISABLE_TRACING turns tracing off entirely.

## Questions answered under this concept

- [Do OpenAI Agents SDK guardrails run on every agent after a handoff?](https://tessary.ai/answers/openai-agents-sdk-evals/do-openai-agents-sdk-guardrails-run-on-every-agent-after-a-handoff)
- [Does the OpenAI Agents SDK support MCP tools?](https://tessary.ai/answers/openai-agents-sdk-evals/does-the-openai-agents-sdk-support-mcp-tools)
- [How do I catch a regression in an Agents SDK workflow?](https://tessary.ai/answers/openai-agents-sdk-evals/how-do-i-catch-a-regression-in-an-agents-sdk-workflow)
- [How do I grade a handoff between two agents?](https://tessary.ai/answers/openai-agents-sdk-evals/how-do-i-grade-a-handoff-between-two-agents)
- [How do I trace an OpenAI Agents SDK run?](https://tessary.ai/answers/openai-agents-sdk-evals/how-do-i-trace-an-openai-agents-sdk-run)
- [Is the OpenAI Agents SDK open source?](https://tessary.ai/answers/openai-agents-sdk-evals/is-the-openai-agents-sdk-open-source)
- [Should a guardrail result be graded separately from the output?](https://tessary.ai/answers/openai-agents-sdk-evals/should-a-guardrail-result-be-graded-separately-from-the-output)
- [What happens when an OpenAI Agents SDK guardrail trips?](https://tessary.ai/answers/openai-agents-sdk-evals/what-happens-when-an-openai-agents-sdk-guardrail-trips)
- [What's the difference between the OpenAI Agents SDK and the Responses API?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-openai-agents-sdk-and-responses-api)
- [What's the difference between the OpenAI Agents SDK and AutoGen?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-autogen)
- [What's the difference between the OpenAI Agents SDK and CrewAI?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-crewai)
- [What's the difference between the OpenAI Agents SDK and Google's Agent Development Kit (ADK)?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-google-adk)
- [What's the difference between the OpenAI Agents SDK and LangChain?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-langchain)
- [What's the difference between the OpenAI Agents SDK and LangGraph?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-langgraph)
- [What's the difference between the OpenAI Agents SDK and Pydantic AI?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-pydantic-ai)
- [What's the difference between the OpenAI Agents SDK and the Codex SDK?](https://tessary.ai/answers/openai-agents-sdk-evals/whats-the-difference-between-the-openai-agents-sdk-and-the-codex-sdk)

---

Source: https://tessary.ai/answers/openai-agents-sdk-evals
All concepts: https://tessary.ai/answers
From Tessary, agent reliability for AI agents in production: https://tessary.ai