From 5f13f101c51dbc5e5f438c2ee8e42e09de967fce Mon Sep 17 00:00:00 2001
From: Virak Ngauv <36271504+virakngauv@users.noreply.github.com>
Date: Thu, 9 Oct 2025 12:23:12 -0700
Subject: [PATCH] Update guardrails.mdx

Update typo that incorrectly explained that output guardrails get the same input as input guardrails (i.e. same as the agent input) instead of from the agent output.
---
 docs/src/content/docs/guides/guardrails.mdx | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/src/content/docs/guides/guardrails.mdx b/docs/src/content/docs/guides/guardrails.mdx
index cf179607..8cfd5d22 100644
--- a/docs/src/content/docs/guides/guardrails.mdx
+++ b/docs/src/content/docs/guides/guardrails.mdx
@@ -27,9 +27,9 @@ Input guardrails run in three steps:
 
 ## Output guardrails
 
-Output guardrails follow the same pattern:
+Output guardrails run in 3 steps:
 
-1. The guardrail receives the same input passed to the agent.
+1. The guardrail receives the output produced by the agent.
 2. The guardrail function executes and returns a [`GuardrailFunctionOutput`](/openai-agents-js/openai/agents/interfaces/guardrailfunctionoutput) wrapped inside an [`OutputGuardrailResult`](/openai-agents-js/openai/agents/interfaces/outputguardrailresult).
 3. If `tripwireTriggered` is `true`, an [`OutputGuardrailTripwireTriggered`](/openai-agents-js/openai/agents/classes/outputguardrailtripwiretriggered) error is thrown.