From 5f886dc73a13cf1798a20f444aefb370797ae5d0 Mon Sep 17 00:00:00 2001
From: Diego Rodrigues de Sa e Souza
 <8016841+diegosouzapw@users.noreply.github.com>
Date: Fri, 29 May 2026 00:58:01 -0300
Subject: [PATCH] fix(translator): strip safety_identifier in openai-responses
 cleanup (#2770) (#2809)

Integrated into release/v3.8.6.
---
 CHANGELOG.md                                    |  1 +
 open-sse/translator/request/openai-responses.ts |  3 +++
 .../translator-openai-responses-req.test.ts     | 17 +++++++++++++++++
 3 files changed, 21 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index b13475ba0f..428e911361 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -22,6 +22,7 @@
 - **combo:** resolve custom `openai-compatible-responses-*` provider targets correctly when called via combo name — combo steps storing the internal UUID-prefixed provider id now match the provider node by id as well as by prefix, fixing 503 errors for users with custom providers used inside combos (#2778)
 - **combos:** fix combo handling so transient 429 rate limit errors do not poison or persist the rate limited state for the same-provider connection (#2800 — thanks @apoapostolov)
 - **gemini:** translate signature-less Gemini thinking model tool calls to text parts to prevent `400 "missing thought_signature"` errors (#2801 — thanks @herjarsa)
+- **translator:** strip `safety_identifier` from `/v1/responses` body before forwarding to Chat Completions upstream; fixes LobeHub-originated `400` errors (#2770)
 - **warning-cleanup:** relax node engine constraint to `>=22.0.0` and clean dependencies (keeping `marked-terminal` to prevent TUI REPL crash) (#2792 — thanks @oyi77)
 - **fix(cli):** register openclaw in the CLI tool-detector so it appears in `omniroute status` alongside its existing API and config support ([#2833](https://github.com/diegosouzapw/OmniRoute/issues/2833))
 
diff --git a/open-sse/translator/request/openai-responses.ts b/open-sse/translator/request/openai-responses.ts
index eadb2e8811..8b25bbb67e 100644
--- a/open-sse/translator/request/openai-responses.ts
+++ b/open-sse/translator/request/openai-responses.ts
@@ -345,6 +345,9 @@ export function openaiResponsesToOpenAIRequest(
     }
   }
   delete result.reasoning;
+  // Strip Responses-API-only fields that Chat Completions rejects with 400.
+  // safety_identifier is sent by LobeHub and has no Chat Completions equivalent (#2770).
+  delete result.safety_identifier;
 
   return result;
 }
diff --git a/tests/unit/translator-openai-responses-req.test.ts b/tests/unit/translator-openai-responses-req.test.ts
index f7e3587a98..ece658aa5d 100644
--- a/tests/unit/translator-openai-responses-req.test.ts
+++ b/tests/unit/translator-openai-responses-req.test.ts
@@ -216,6 +216,23 @@ test("Responses -> Chat passes through when background flag is unset or false (n
   }
 });
 
+test("Responses -> Chat strips safety_identifier (LobeHub #2770)", () => {
+  // LobeHub sends safety_identifier in Responses API bodies. Chat Completions rejects it
+  // with HTTP 400. The translator must strip it in the Responses-API cleanup block.
+  const result = openaiResponsesToOpenAIRequest(
+    "gpt-4o",
+    {
+      input: [{ role: "user", content: [{ type: "input_text", text: "hi" }] }],
+      safety_identifier: "sid-xyz",
+    },
+    false,
+    null
+  ) as Record<string, unknown>;
+
+  assert.equal(result.safety_identifier, undefined, "safety_identifier must be stripped before forwarding to Chat Completions");
+  assert.ok(Array.isArray(result.messages), "translation must still produce messages");
+});
+
 test("Chat -> Responses converts messages, tool calls, tool outputs, tools and pass-through params", () => {
   const result = openaiToOpenAIResponsesRequest(
     "gpt-4o",