From 3a646ea0a94f04d6077a93520d2cd38c5966ed57 Mon Sep 17 00:00:00 2001
From: ghost <49853598+JSONbored@users.noreply.github.com>
Date: Sat, 13 Jun 2026 00:55:22 -0700
Subject: [PATCH] fix(submissions): block cross-category website duplicates

---
 apps/submission-gate/src/duplicates.ts |  2 ++
 tests/submission-gate-worker.test.ts   | 46 ++++++++++++++++++++++++++
 2 files changed, 48 insertions(+)

diff --git a/apps/submission-gate/src/duplicates.ts b/apps/submission-gate/src/duplicates.ts
index 16aebddeb4..ae35cf7f33 100644
--- a/apps/submission-gate/src/duplicates.ts
+++ b/apps/submission-gate/src/duplicates.ts
@@ -83,12 +83,14 @@ const CROSS_CATEGORY_STRICT_URL_FIELDS = new Set([
   "repoUrl",
   "repositoryUrl",
   "sourceUrl",
+  "websiteUrl",
   "download_url",
   "github_url",
   "package_url",
   "repo_url",
   "repository_url",
   "source_url",
+  "website_url",
 ]);
 const DOMAIN_ONLY_EXCLUSIONS = new Set([
   "github.com",
diff --git a/tests/submission-gate-worker.test.ts b/tests/submission-gate-worker.test.ts
index 5fb1c3db50..d474aa5262 100644
--- a/tests/submission-gate-worker.test.ts
+++ b/tests/submission-gate-worker.test.ts
@@ -2762,6 +2762,52 @@ repoUrl: "https://github.com/langchain-ai/langchain.git"
     );
   });
 
+  it("treats same canonical website across different categories as a strict duplicate", () => {
+    const existingTool = extractContentDuplicateSignals({
+      filePath: "content/tools/acme-claude.mdx",
+      content: `---
+title: Acme Claude
+slug: acme-claude
+category: tools
+description: Tooling for Acme Claude workflows.
+websiteUrl: "https://acme-claude.example/product"
+---
+`,
+    });
+    const candidateMcp = extractContentDuplicateSignals({
+      filePath: "content/mcp/acme-claude-server.mdx",
+      content: `---
+title: Acme Claude MCP Server
+slug: acme-claude-server
+category: mcp
+description: MCP server for Acme Claude workflows.
+websiteUrl: "https://acme-claude.example/product?utm_source=submission"
+---
+`,
+    });
+
+    expect(
+      findStrictContentDuplicateMatch(candidateMcp, [existingTool]),
+    ).toMatchObject({
+      reasons: expect.arrayContaining([
+        expect.stringContaining(
+          "same canonical source URL https://acme-claude.example/product across mcp/tools",
+        ),
+      ]),
+    });
+    expect(findRelatedContentMatches(candidateMcp, [existingTool])).toEqual(
+      expect.arrayContaining([
+        expect.objectContaining({
+          reasons: expect.arrayContaining([
+            expect.stringContaining(
+              "same canonical source URL https://acme-claude.example/product across mcp/tools",
+            ),
+          ]),
+        }),
+      ]),
+    );
+  });
+
   it("treats collection member overlap as related context, not a strict duplicate", () => {
     const existingTool = extractContentDuplicateSignals({
       filePath: "content/tools/storybook-a11y.mdx",