fix(codex): strip residual Chat Completions fields for gpt-5.5 Responses API (#2608)

open-sse/executors/codex.ts

@@ -1286,21 +1286,35 @@ export class CodexExecutor extends BaseExecutor {
       return body;
     }
 
-    // Remove unsupported parameters for Codex API
-    delete body.temperature;
-    delete body.top_p;
-    delete body.frequency_penalty;
-    delete body.presence_penalty;
-    delete body.logprobs;
-    delete body.top_logprobs;
-    delete body.n;
-    delete body.seed;
-    // max_tokens and max_output_tokens already deleted above (before passthrough return)
-    delete body.user; // Cursor sends this but Codex doesn't support it
-
-    delete body.metadata; // Cursor sends this but Codex doesn't support it
-    delete body.stream_options; // Cursor sends this but Codex doesn't support it
-    delete body.safety_identifier; // Droid CLI sends this but Codex doesn't support it
+    // Issue #2608: Use an allowlist of known Responses API fields instead of a
+    // denylist of Chat Completions fields. The denylist approach missed fields
+    // like `stop`, `response_format`, `logit_bias`, `function_call`, `functions`,
+    // `max_completion_tokens`, and `parallel_tool_calls` — causing gpt-5.5 to
+    // reject with "routing_unsupported" (400). An allowlist is future-proof:
+    // any unknown field from Chat Completions (or other formats) is stripped.
+    const RESPONSES_API_ALLOWLIST = new Set([
+      "model",
+      "input",
+      "instructions",
+      "tools",
+      "tool_choice",
+      "stream",
+      "store",
+      "reasoning",
+      "service_tier",
+      "include",
+      "previous_response_id",
+      "prompt_cache_key",
+      "client_metadata",
+      // Internal markers used by OmniRoute pipeline
+      "_omnirouteResponsesStore",
+    ]);
+
+    for (const key of Object.keys(body)) {
+      if (!RESPONSES_API_ALLOWLIST.has(key)) {
+        delete body[key];
+      }
+    }
 
     return body;
   }

tests/unit/executor-codex.test.ts

@@ -216,6 +216,76 @@ test("CodexExecutor.transformRequest injects default instructions, clamps reason
   assert.equal(result.stream_options, undefined);
 });
 
+// Issue #2608: gpt-5.5 models reject residual Chat Completions fields via Codex OAuth.
+// The non-passthrough path must strip ALL non-Responses-API fields using an allowlist.
+test("CodexExecutor.transformRequest non-passthrough allowlist strips all residual Chat Completions fields (#2608)", () => {
+  const executor = new CodexExecutor();
+  const body = {
+    model: "gpt-5.5",
+    messages: [{ role: "user", content: "hello" }],
+    instructions: "",
+    // All of these are Chat Completions fields that must be stripped:
+    temperature: 0.7,
+    top_p: 0.9,
+    frequency_penalty: 0.5,
+    presence_penalty: 0.3,
+    logprobs: true,
+    top_logprobs: 3,
+    n: 2,
+    seed: 42,
+    stop: ["\n"],
+    response_format: { type: "json_object" },
+    logit_bias: { "123": 1 },
+    function_call: "auto",
+    functions: [{ name: "test", parameters: {} }],
+    max_completion_tokens: 1000,
+    parallel_tool_calls: true,
+    user: "cursor-user",
+    metadata: { key: "value" },
+    stream_options: { include_usage: true },
+    safety_identifier: "safe-1",
+    suffix: "end",
+    // Custom/arbitrary fields that could be injected by middleware
+    custom_field: "should be stripped",
+    _internal_marker: true,
+  };
+
+  const result = executor.transformRequest("gpt-5.5", body, false, {
+    requestEndpointPath: "/responses",
+  });
+
+  // Allowed Responses API fields should survive
+  assert.equal(result.model, "gpt-5.5");
+  assert.ok(Array.isArray(result.input));
+  assert.equal(typeof result.instructions, "string");
+  assert.equal(result.store, false);
+  assert.equal(result.stream, true);
+
+  // All Chat Completions fields must be stripped
+  assert.equal(result.temperature, undefined, "temperature should be stripped");
+  assert.equal(result.top_p, undefined, "top_p should be stripped");
+  assert.equal(result.frequency_penalty, undefined, "frequency_penalty should be stripped");
+  assert.equal(result.presence_penalty, undefined, "presence_penalty should be stripped");
+  assert.equal(result.logprobs, undefined, "logprobs should be stripped");
+  assert.equal(result.top_logprobs, undefined, "top_logprobs should be stripped");
+  assert.equal(result.n, undefined, "n should be stripped");
+  assert.equal(result.seed, undefined, "seed should be stripped");
+  assert.equal(result.stop, undefined, "stop should be stripped");
+  assert.equal(result.response_format, undefined, "response_format should be stripped");
+  assert.equal(result.logit_bias, undefined, "logit_bias should be stripped");
+  assert.equal(result.function_call, undefined, "function_call should be stripped");
+  assert.equal(result.functions, undefined, "functions should be stripped");
+  assert.equal(result.max_completion_tokens, undefined, "max_completion_tokens should be stripped");
+  assert.equal(result.parallel_tool_calls, undefined, "parallel_tool_calls should be stripped");
+  assert.equal(result.user, undefined, "user should be stripped");
+  assert.equal(result.metadata, undefined, "metadata should be stripped");
+  assert.equal(result.stream_options, undefined, "stream_options should be stripped");
+  assert.equal(result.safety_identifier, undefined, "safety_identifier should be stripped");
+  assert.equal(result.suffix, undefined, "suffix should be stripped");
+  assert.equal(result.custom_field, undefined, "arbitrary custom fields should be stripped");
+  assert.equal(result._internal_marker, undefined, "internal markers should be stripped");
+});
+
 test("CodexExecutor.transformRequest normalizes max reasoning_effort to xhigh", () => {
   const executor = new CodexExecutor();
   const result = executor.transformRequest(