From 6265dca1ee62bdbd8b02ef787452d7422773f56b Mon Sep 17 00:00:00 2001
From: "github-actions[bot]"
 <41898282+github-actions[bot]@users.noreply.github.com>
Date: Sun, 30 Aug 2026 03:31:09 +0000
Subject: [PATCH] docs: add Codex auto model caveat to
 authentication-architecture example

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
---
 docs/authentication-architecture.md | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/docs/authentication-architecture.md b/docs/authentication-architecture.md
index 3c2a73607..78126d44f 100644
--- a/docs/authentication-architecture.md
+++ b/docs/authentication-architecture.md
@@ -481,9 +481,15 @@ export OPENAI_API_KEY="sk-..."
 
 sudo awf \
     --allow-domains api.openai.com \
-    "codex --prompt 'Hello world'"
+    "codex --model gpt-5.3-codex --prompt 'Hello world'"
 ```
 
+Use an explicit model such as `gpt-5.3-codex` rather than Codex's `auto` model
+alias. `auto` depends on ChatGPT-authenticated remote model/plugin metadata
+resolution, which API-key based proxy routing cannot use even when
+`chatgpt.com` is allowed, so `auto` can fail with `The requested model is not
+supported` (see [docs/api-proxy-sidecar.md](api-proxy-sidecar.md#codex-openai-example)).
+
 **Example 3: Using both providers**
 
 ```bash