src/app/api/ai-event/route.ts

@@ -2,37 +2,12 @@ import { NextResponse } from "next/server";
 import { auth } from "@/auth";
 import { headers } from "next/headers";
 import { openRouterClient } from "@/lib/openrouter-client";
+import { AiEventRequestSchema, AiEventResponseSchema } from "@/lib/types";
 
-export async function POST(request: Request) {
-	const session = await auth.api.getSession({
-		headers: await headers(),
-	});
+const MODEL = "openai/gpt-5.4-mini";
 
-	if (!session?.user) {
-		return NextResponse.json(
-			{ error: "Authentication required" },
-			{ status: 401 },
-		);
-	}
-
-	const { prompt } = await request.json();
-
-	// Validate prompt input
-	if (!prompt || typeof prompt !== "string" || prompt.trim().length === 0) {
-		return NextResponse.json(
-			{ error: "Prompt is required and must be a non-empty string" },
-			{ status: 400 },
-		);
-	}
-	if (prompt.length > 2000) {
-		return NextResponse.json(
-			{ error: "Prompt must be less than 2000 characters" },
-			{ status: 400 },
-		);
-	}
-
-	const systemPrompt = `
-You are an assistant that converts natural language into an ARRAY of calendar events.
+const buildSystemPrompt = () => `
+You are an assistant that converts natural language and images into an ARRAY of calendar events.
 TypeScript type:
 
 {
@@ -55,25 +30,146 @@ Rules:
   - If no end time is given (and event is not allDay), default to 1 hour after start.
   - If multiple events are described, return multiple.
   - If recurrence is implied (e.g. "every Monday", "daily for 10 days", "monthly on the 15th"), generate a recurrenceRule.
+  - When analyzing an image, extract ALL visible event details: titles, dates, times, locations, descriptions.
   - Output ONLY valid JSON (no prose).
 `;
 
-	try {
+const callTextOnly = async (systemPrompt: string, prompt: string) => {
 	const result = openRouterClient.callModel({
-			model: "openai/gpt-5.4-mini",
+		model: MODEL,
 		instructions: systemPrompt,
 		input: prompt,
 	});
 
-		const text = await result.getText();
-		const parsed = JSON.parse(text);
-		return NextResponse.json(parsed);
+	const rawResponse = await result.getText();
+	return { rawResponse, startTime: performance.now() };
+};
+
+const callMultimodal = async (
+	systemPrompt: string,
+	prompt: string | undefined,
+	imageBase64: string,
+) => {
+	const messages = [
+		{
+			role: "system" as const,
+			content: systemPrompt,
+		},
+		{
+			role: "user" as const,
+			content: [
+				{
+					type: "text" as const,
+					text: prompt || "Extract all calendar events from this image.",
+				},
+				{
+					type: "image_url" as const,
+					imageUrl: { url: imageBase64 },
+				},
+			],
+		},
+	];
+
+	const startTime = performance.now();
+
+	const response = await openRouterClient.chat.send({
+		chatRequest: {
+			model: MODEL,
+			messages,
+		},
+	});
+
+	const rawResponse =
+		typeof response === "object" &&
+		"choices" in response &&
+		response.choices?.[0]?.message
+			? typeof response.choices[0].message.content === "string"
+				? response.choices[0].message.content
+				: JSON.stringify(response.choices[0].message.content)
+			: JSON.stringify(response);
+
+	return { rawResponse, startTime };
+};
+
+const extractJsonFromText = (text: string): unknown => {
+	try {
+		return JSON.parse(text);
+	} catch {
+		const codeBlockMatch = text.match(/```(?:json)?\s*([\s\S]*?)```/);
+		if (codeBlockMatch) {
+			return JSON.parse(codeBlockMatch[1].trim());
+		}
+		const arrayMatch = text.match(/\[[\s\S]*\]/);
+		if (arrayMatch) {
+			return JSON.parse(arrayMatch[0]);
+		}
+		throw new Error(`No JSON found in response: ${text.slice(0, 200)}`);
+	}
+};
+
+export async function POST(request: Request) {
+	const session = await auth.api.getSession({
+		headers: await headers(),
+	});
+
+	if (!session?.user) {
+		return NextResponse.json(
+			{ error: "Authentication required" },
+			{ status: 401 },
+		);
+	}
+
+	const body = await request.json();
+	const parsedInput = AiEventRequestSchema.safeParse(body);
+
+	if (!parsedInput.success) {
+		return NextResponse.json(
+			{
+				error: "Invalid input",
+				details: parsedInput.error.flatten().fieldErrors,
+			},
+			{ status: 400 },
+		);
+	}
+
+	const { prompt, imageBase64 } = parsedInput.data;
+	const inputMode = imageBase64 ? "multimodal" : "text";
+	const systemPrompt = buildSystemPrompt();
+	let rawResponse: string | undefined;
+
+	try {
+		const result =
+			inputMode === "multimodal"
+				? await callMultimodal(systemPrompt, prompt, imageBase64!)
+				: await callTextOnly(systemPrompt, prompt!);
+
+		rawResponse = result.rawResponse;
+
+		const rawJson = extractJsonFromText(rawResponse);
+		const validated = AiEventResponseSchema.safeParse(rawJson);
+
+		if (!validated.success) {
+			console.error("AI response validation failed:", {
+				issues: validated.error.flatten().fieldErrors,
+				rawResponse,
+			});
+
+			return NextResponse.json(
+				{
+					error: "AI returned invalid event data",
+					details: validated.error.flatten().fieldErrors,
+				},
+				{ status: 422 },
+			);
+		}
+
+		return NextResponse.json(validated.data);
 	} catch (error) {
 		console.error("AI Event Creation Error:", error);
 		return NextResponse.json(
 			{
 				error: "Failed to parse AI output",
-				raw: error instanceof Error ? error.message : error,
+				raw: error instanceof Error ? error.message : String(error),
 			},
 			{ status: 500 },
 		);