feat(ai): standardized codex reasoning context to all_turns

- Standardized reasoning context to default to all_turns across all Codex requests to mirror codex-rs behavior.
- Updated request transformer logic to explicitly set reasoning context rather than omitting it when unset.
- Updated API documentation to reflect consistent all_turns default behavior and corrected internal references.
- Adjusted unit tests to verify the enforced all_turns default reasoning context expectation.

packages/ai/CHANGELOG.md

@@ -2,6 +2,10 @@
 
 ## [Unreleased]
 
+### Changed
+
+- Default reasoning context to `all_turns` for all Codex requests
+
 ## [16.2.2] - 2026-06-27
 
 ### Added

packages/ai/src/providers/openai-codex-responses.ts

@@ -104,7 +104,7 @@ import { transformMessages } from "./transform-messages";
 export interface OpenAICodexResponsesOptions extends StreamOptions {
 	reasoning?: "none" | "minimal" | "low" | "medium" | "high" | "xhigh";
 	reasoningSummary?: "auto" | "concise" | "detailed" | null;
-	/** `reasoning.context` replay scope. Defaults to `all_turns` under {@link OpenAICodexResponsesOptions.responsesLite}, otherwise omitted (server default is `current_turn`). */
+	/** `reasoning.context` replay scope; defaults to `all_turns` for every Codex request when unset. */
 	reasoningContext?: CodexReasoningContext;
 	textVerbosity?: "low" | "medium" | "high";
 	include?: string[];
@@ -117,7 +117,7 @@ export interface OpenAICodexResponsesOptions extends StreamOptions {
 	 * `x-openai-internal-codex-responses-lite: true` on HTTP requests and on the
 	 * WebSocket upgrade (the marker is connection-scoped there, so lite and
 	 * non-lite turns never share a pooled socket), strips image detail from
-	 * input, and defaults `reasoning.context` to `all_turns` — mirroring codex-rs.
+	 * input, and disables parallel tool calling — mirroring codex-rs.
 	 */
 	responsesLite?: boolean;
 	/**

packages/ai/src/providers/openai-codex/request-transformer.ts

@@ -14,11 +14,11 @@ export interface ReasoningConfig {
 export interface CodexRequestOptions {
 	reasoningEffort?: ReasoningConfig["effort"];
 	reasoningSummary?: ReasoningConfig["summary"] | null;
-	/** Explicit `reasoning.context` override. Defaults to `all_turns` under {@link CodexRequestOptions.responsesLite}, otherwise omitted (server default is `current_turn`). */
+	/** Explicit `reasoning.context` override; defaults to `all_turns` for every Codex request when unset. */
 	reasoningContext?: CodexReasoningContext;
 	textVerbosity?: "low" | "medium" | "high";
 	include?: string[];
-	/** Responses Lite transport contract: strips image detail and defaults `reasoning.context` to `all_turns`, mirroring codex-rs. */
+	/** Responses Lite transport contract: strips image detail and disables parallel tool calling, mirroring codex-rs. */
 	responsesLite?: boolean;
 }
 
@@ -254,13 +254,9 @@ export async function transformRequestBody(
 			...body.reasoning,
 			...reasoningConfig,
 		};
-		// Responses Lite keeps reasoning replay server-side; codex-rs requests
-		// `all_turns` there and otherwise omits context so the server default
-		// (currently `current_turn`) applies.
-		const reasoningContext = options.reasoningContext ?? "all_turns";
-		if (reasoningContext !== undefined) {
-			body.reasoning.context = reasoningContext;
-		}
+		// Default reasoning replay to `all_turns` for every Codex request,
+		// mirroring codex-rs; an explicit `reasoningContext` overrides it.
+		body.reasoning.context = options.reasoningContext ?? "all_turns";
 	} else {
 		delete body.reasoning;
 	}

packages/ai/test/openai-codex-responses-lite.test.ts

@@ -83,7 +83,7 @@ function createCodexFetchMock(sse: string, onRequest: (captured: CapturedCodexRe
 }
 
 describe("openai-codex reasoning.context", () => {
-	it("forwards an explicit reasoning.context and omits it by default", async () => {
+	it("forwards an explicit reasoning.context and defaults to all_turns", async () => {
 		const model = createCodexModel("gpt-5.1-codex");
 
 		const explicit = await transformRequestBody({ model: model.id }, model, {
@@ -92,8 +92,8 @@ describe("openai-codex reasoning.context", () => {
 		});
 		expect(explicit.reasoning?.context).toBe("current_turn");
 
-		const omitted = await transformRequestBody({ model: model.id }, model, { reasoningEffort: "medium" });
-		expect(omitted.reasoning?.context).toBeUndefined();
+		const defaulted = await transformRequestBody({ model: model.id }, model, { reasoningEffort: "medium" });
+		expect(defaulted.reasoning?.context).toBe("all_turns");
 	});
 
 	it("defaults reasoning.context to all_turns under Responses Lite unless overridden", async () => {