From 1b3d1caa2180d620d478bad7561f1ce6d576d417 Mon Sep 17 00:00:00 2001
From: anirudh5harma <anirudh1304@gmail.com>
Date: Sun, 28 Jun 2026 02:53:54 +0530
Subject: [PATCH] Fix social URL host matching

---
 apps/web/lib/url-helpers.test.ts | 59 ++++++++++++++++++++++++++++++++
 apps/web/lib/url-helpers.ts      | 41 ++++++++++++++++++----
 2 files changed, 93 insertions(+), 7 deletions(-)
 create mode 100644 apps/web/lib/url-helpers.test.ts

diff --git a/apps/web/lib/url-helpers.test.ts b/apps/web/lib/url-helpers.test.ts
new file mode 100644
index 000000000..8f6218682
--- /dev/null
+++ b/apps/web/lib/url-helpers.test.ts
@@ -0,0 +1,59 @@
+import { describe, expect, it } from "bun:test"
+import {
+	collectValidUrls,
+	isLinkedInProfileUrl,
+	isTwitterUrl,
+} from "./url-helpers"
+
+describe("social URL detection", () => {
+	it("matches Twitter and X by hostname only", () => {
+		expect(isTwitterUrl("https://twitter.com/supermemoryai")).toBe(true)
+		expect(isTwitterUrl("https://mobile.twitter.com/supermemoryai")).toBe(true)
+		expect(isTwitterUrl("https://x.com/supermemoryai")).toBe(true)
+		expect(isTwitterUrl("x.com/supermemoryai")).toBe(true)
+
+		expect(isTwitterUrl("https://exampletwitter.com/supermemoryai")).toBe(false)
+		expect(isTwitterUrl("https://notx.com/profile")).toBe(false)
+		expect(isTwitterUrl("https://example.com/twitter.com/supermemoryai")).toBe(
+			false,
+		)
+		expect(isTwitterUrl("https://x.com.evil.example/supermemoryai")).toBe(false)
+	})
+
+	it("matches LinkedIn profile URLs by hostname and /in/ path", () => {
+		expect(isLinkedInProfileUrl("https://linkedin.com/in/supermemoryai")).toBe(
+			true,
+		)
+		expect(
+			isLinkedInProfileUrl("https://www.linkedin.com/in/supermemoryai"),
+		).toBe(true)
+		expect(isLinkedInProfileUrl("linkedin.com/in/supermemoryai")).toBe(true)
+
+		expect(
+			isLinkedInProfileUrl(
+				"https://evil.example/linkedin.com/in/supermemoryai",
+			),
+		).toBe(false)
+		expect(isLinkedInProfileUrl("https://linkedin.com/company/acme")).toBe(
+			false,
+		)
+		expect(isLinkedInProfileUrl("https://notlinkedin.com/in/person")).toBe(
+			false,
+		)
+		expect(isLinkedInProfileUrl("https://linkedin.com/in")).toBe(false)
+	})
+})
+
+describe("collectValidUrls", () => {
+	it("keeps non-Twitter domains that only contain x.com as text", () => {
+		expect(collectValidUrls("", ["https://notx.com/profile"])).toEqual([
+			"https://notx.com/profile",
+		])
+	})
+
+	it("rejects LinkedIn profile URLs on unrelated hosts", () => {
+		expect(
+			collectValidUrls("https://evil.example/linkedin.com/in/person", []),
+		).toEqual([])
+	})
+})
diff --git a/apps/web/lib/url-helpers.ts b/apps/web/lib/url-helpers.ts
index 0d3afa6ae..741f746a6 100644
--- a/apps/web/lib/url-helpers.ts
+++ b/apps/web/lib/url-helpers.ts
@@ -120,13 +120,40 @@ export const extractUrls = (
 	return { urls, duplicates }
 }
 
+const parseWebUrl = (url: string): URL | null => {
+	const trimmed = url.trim()
+	if (!trimmed) return null
+
+	try {
+		const parsed = new URL(trimmed)
+		return parsed.protocol === "http:" || parsed.protocol === "https:"
+			? parsed
+			: null
+	} catch {
+		try {
+			return new URL(`https://${trimmed}`)
+		} catch {
+			return null
+		}
+	}
+}
+
+const hostnameMatches = (hostname: string, domain: string): boolean => {
+	const normalizedHostname = hostname.toLowerCase()
+	return (
+		normalizedHostname === domain || normalizedHostname.endsWith(`.${domain}`)
+	)
+}
+
 /**
  * Checks if a URL is a Twitter/X URL.
  */
 export const isTwitterUrl = (url: string): boolean => {
-	const normalizedUrl = url.toLowerCase()
+	const parsed = parseWebUrl(url)
+	if (!parsed) return false
 	return (
-		normalizedUrl.includes("twitter.com") || normalizedUrl.includes("x.com")
+		hostnameMatches(parsed.hostname, "twitter.com") ||
+		hostnameMatches(parsed.hostname, "x.com")
 	)
 }
 
@@ -134,11 +161,11 @@ export const isTwitterUrl = (url: string): boolean => {
  * Checks if a URL is a LinkedIn profile URL (not a company page).
  */
 export const isLinkedInProfileUrl = (url: string): boolean => {
-	const normalizedUrl = url.toLowerCase()
-	return (
-		normalizedUrl.includes("linkedin.com/in/") &&
-		!normalizedUrl.includes("linkedin.com/company/")
-	)
+	const parsed = parseWebUrl(url)
+	if (!parsed || !hostnameMatches(parsed.hostname, "linkedin.com")) return false
+
+	const [section, handle] = parsed.pathname.split("/").filter(Boolean)
+	return section?.toLowerCase() === "in" && Boolean(handle)
 }
 
 /**