From 74461e6f53d3a8a99c5e1e4b4a352af330665a1a Mon Sep 17 00:00:00 2001
From: roboomp <omp@can.ac>
Date: Sun, 17 May 2026 10:55:29 +0000
Subject: [PATCH] fix(ai): broadened openai schema sanitization coverage

Added dependencies and contentSchema to the OpenAI Responses schema-position sets so draft-04..07 dependencies maps and draft 2019-09 contentSchema nodes also get the properties-on-object normalization. Recognized array-form type declarations that include object. Preserved malformed non-array oneOf payloads instead of dropping them. Documented the cycle-safe cache seed and added a self-referential regression test.

Refs #1147
---
 packages/ai/src/utils/schema/normalize.ts     | 37 +++++++++++-
 packages/ai/test/schema-normalization.test.ts | 57 +++++++++++++++++++
 2 files changed, 92 insertions(+), 2 deletions(-)

diff --git a/packages/ai/src/utils/schema/normalize.ts b/packages/ai/src/utils/schema/normalize.ts
index 689c5ab3b..befaf81a0 100644
--- a/packages/ai/src/utils/schema/normalize.ts
+++ b/packages/ai/src/utils/schema/normalize.ts
@@ -857,6 +857,11 @@ const OPENAI_RESPONSES_SCHEMA_ARRAY_KEYS = new Set(["anyOf", "oneOf", "allOf", "
 const OPENAI_RESPONSES_SCHEMA_MAP_KEYS = new Set([
 	"properties",
 	"patternProperties",
+	// `dependencies` is the Draft-04..07 schema-valued form; older MCP servers
+	// still emit `{ dependencies: { foo: { type: "object" } } }`. String-array
+	// branches per key pass through `normalizeOpenAIResponsesSchemaNode`
+	// untouched because non-objects return as-is.
+	"dependencies",
 	"dependentSchemas",
 	"$defs",
 	"definitions",
@@ -865,6 +870,7 @@ const OPENAI_RESPONSES_SCHEMA_VALUE_KEYS = new Set([
 	"items",
 	"additionalItems",
 	"contains",
+	"contentSchema",
 	"propertyNames",
 	"if",
 	"then",
@@ -903,13 +909,24 @@ function normalizeOpenAIResponsesSchemaNode(value: unknown, cache: WeakMap<JsonO
 	const cached = cache.get(value);
 	if (cached) return cached;
 
+	// Seed the cache with the in-flight `output` BEFORE recursing so that a
+	// child re-entering this node mid-walk gets the partial back instead of
+	// triggering an infinite recursion. A cycle hitting this seeded entry
+	// forces `changed = true` below (the cached partial is referentially
+	// distinct from `value`), which is why the final `cache.set(value, result)`
+	// never silently overwrites the seed with `value` on a cyclic input.
 	const output: JsonObject = {};
 	cache.set(value, output);
 
 	let changed = false;
 	for (const key in value) {
 		if (!Object.hasOwn(value, key)) continue;
-		if (key === "oneOf") {
+		// Drop only well-formed `oneOf` arrays here; they are re-emitted as
+		// `anyOf` after the loop so any neighboring `anyOf` entries can be
+		// concatenated. A non-array `oneOf` is malformed for the wire but
+		// still preserved verbatim so callers can see the original payload
+		// instead of having it silently disappear.
+		if (key === "oneOf" && Array.isArray(value.oneOf)) {
 			changed = true;
 			continue;
 		}
@@ -936,16 +953,32 @@ function normalizeOpenAIResponsesSchemaNode(value: unknown, cache: WeakMap<JsonO
 			: rewrittenOneOf;
 	}
 
-	if (value.type === "object" && !Object.hasOwn(value, "properties")) {
+	// Draft 2020-12 lets `type` be an array (e.g. `["object", "null"]`); treat
+	// any variant that includes "object" as an object position for the
+	// properties requirement.
+	if (declaresObjectType(value.type) && !Object.hasOwn(value, "properties")) {
 		output.properties = {};
 		changed = true;
 	}
 
+	// Safe to overwrite the seed: any cyclic re-entry above already observed
+	// the seeded partial and set `changed = true` for that node, so a node
+	// that finishes with `changed === false` is provably non-cyclic and
+	// referentially equal to its input.
 	const result = changed ? output : value;
 	cache.set(value, result);
 	return result;
 }
 
+function declaresObjectType(type: unknown): boolean {
+	if (type === "object") return true;
+	if (!Array.isArray(type)) return false;
+	for (const variant of type) {
+		if (variant === "object") return true;
+	}
+	return false;
+}
+
 function normalizeOpenAIResponsesSchemaArray(value: unknown[], cache: WeakMap<JsonObject, JsonObject>): unknown[] {
 	let changed = false;
 	const output = value.map(item => {
diff --git a/packages/ai/test/schema-normalization.test.ts b/packages/ai/test/schema-normalization.test.ts
index c1a0a6657..a1e871b81 100644
--- a/packages/ai/test/schema-normalization.test.ts
+++ b/packages/ai/test/schema-normalization.test.ts
@@ -387,6 +387,63 @@ describe("sanitizeSchemaForOpenAIResponses", () => {
 			anyOf: [{ type: "object", properties: {} }],
 		});
 	});
+
+	it("adds empty properties under draft-07 dependencies and draft 2019-09 contentSchema", () => {
+		const schema = {
+			type: "object",
+			properties: {
+				body: {
+					type: "string",
+					contentSchema: { type: "object" },
+				},
+			},
+			dependencies: {
+				body: { type: "object" },
+				other: ["body"],
+			},
+		};
+
+		expect(sanitizeSchemaForOpenAIResponses(schema)).toEqual({
+			type: "object",
+			properties: {
+				body: {
+					type: "string",
+					contentSchema: { type: "object", properties: {} },
+				},
+			},
+			dependencies: {
+				body: { type: "object", properties: {} },
+				other: ["body"],
+			},
+		});
+	});
+
+	it("adds empty properties when `type` is a draft 2020-12 array including object", () => {
+		expect(sanitizeSchemaForOpenAIResponses({ type: ["object", "null"] })).toEqual({
+			type: ["object", "null"],
+			properties: {},
+		});
+	});
+
+	it("preserves non-array oneOf payloads verbatim instead of dropping them", () => {
+		const malformed = { type: "object", oneOf: { type: "object" } } as unknown as Record<string, unknown>;
+
+		expect(sanitizeSchemaForOpenAIResponses(malformed)).toEqual({
+			type: "object",
+			oneOf: { type: "object" },
+			properties: {},
+		});
+	});
+
+	it("does not recurse infinitely on self-referential object schemas", () => {
+		const circular: Record<string, unknown> = { type: "object", properties: {} };
+		(circular.properties as Record<string, unknown>).self = circular;
+
+		const sanitized = sanitizeSchemaForOpenAIResponses(circular);
+		const properties = (sanitized as { properties: Record<string, unknown> }).properties;
+		expect(properties.self).toBe(sanitized as unknown as object);
+		expect((sanitized as { type: unknown }).type).toBe("object");
+	});
 });
 
 // ---------------------------------------------------------------------------