feat(vom): vom fix

apps/extension/src/tools/__tests__/observation.test.ts

@@ -424,6 +424,73 @@ describe("buildVomScene", () => {
     );
   });
 
+  it("annotates external links from the explicit page URL without reading window.location", () => {
+    const axNodes: CdpAxNode[] = [
+      {
+        nodeId: "1",
+        role: { type: "role", value: "RootWebArea" },
+        backendDOMNodeId: 100,
+        childIds: ["2", "3"],
+      },
+      {
+        nodeId: "2",
+        parentId: "1",
+        role: { type: "role", value: "Link" },
+        name: { type: "computedString", value: "Docs" },
+        backendDOMNodeId: 200,
+      },
+      {
+        nodeId: "3",
+        parentId: "1",
+        role: { type: "role", value: "link" },
+        name: { type: "computedString", value: "Local" },
+        backendDOMNodeId: 300,
+      },
+    ];
+    const captured: CapturedViewModel = {
+      viewport: { width: 1000, height: 800 },
+      iframeNodes: new Map(),
+      excludedBackendNodeIds: new Set(),
+      nodes: [
+        {
+          backendNodeId: 100,
+          parentBackendNodeId: null,
+          tag: "body",
+          attrs: {},
+          rect: { x: 0, y: 0, w: 1000, h: 800 },
+          paintOrder: 0,
+          position: "static",
+          pointerEvents: "auto",
+        },
+        {
+          backendNodeId: 200,
+          parentBackendNodeId: 100,
+          tag: "a",
+          attrs: { href: "https://docs.example.org/start" },
+          rect: { x: 10, y: 10, w: 60, h: 20 },
+          paintOrder: 1,
+          position: "static",
+          pointerEvents: "auto",
+        },
+        {
+          backendNodeId: 300,
+          parentBackendNodeId: 100,
+          tag: "a",
+          attrs: { href: "/local" },
+          rect: { x: 80, y: 10, w: 60, h: 20 },
+          paintOrder: 2,
+          position: "static",
+          pointerEvents: "auto",
+        },
+      ],
+    };
+
+    const scene = buildVomScene(axNodes, captured, { pageUrl: "https://app.example.test/home" });
+
+    expect(scene.nodes.find((node) => node.id === 200)?.href).toBe("docs.example.org");
+    expect(scene.nodes.find((node) => node.id === 300)?.href).toBeUndefined();
+  });
+
   it("uses the nearest backend AX ancestor as parentId", () => {
     const axNodes: CdpAxNode[] = [
       {

apps/extension/src/tools/observation.ts

@@ -969,7 +969,29 @@ function axNodeInOverlaySubtree(
   );
 }
 
-export function buildVomScene(axNodes: CdpAxNode[], captured: CapturedViewModel): VomScene {
+export interface BuildVomSceneOptions {
+  pageUrl?: string;
+}
+
+function externalHrefHost(
+  href: string | undefined,
+  pageUrl: string | undefined,
+): string | undefined {
+  if (!href || !pageUrl) return undefined;
+  try {
+    const page = new URL(pageUrl);
+    const target = new URL(href, page);
+    return target.origin !== page.origin ? target.hostname : undefined;
+  } catch {
+    return undefined;
+  }
+}
+
+export function buildVomScene(
+  axNodes: CdpAxNode[],
+  captured: CapturedViewModel,
+  options: BuildVomSceneOptions = {},
+): VomScene {
   const signals = capturedOnlySignals(captured.nodes);
   const { capturedByBackendId } = signals;
   const axSignals = buildAxSignals(axNodes);
@@ -993,15 +1015,9 @@ export function buildVomScene(axNodes: CdpAxNode[], captured: CapturedViewModel)
 
     // For link nodes, attach external hostname so the renderer can annotate it.
     // Relative hrefs and same-origin hrefs are omitted to avoid noise.
-    if (role === "link" && capturedNode?.attrs.href) {
-      try {
-        const target = new URL(capturedNode.attrs.href, window.location.href);
-        if (target.origin !== window.location.origin) {
-          vomNode.href = target.hostname;
-        }
-      } catch {
-        // Relative or non-URL href (e.g. javascript:) — skip
-      }
+    if (role?.toLowerCase() === "link") {
+      const hrefHost = externalHrefHost(capturedNode?.attrs.href, options.pageUrl);
+      if (hrefHost) vomNode.href = hrefHost;
     }
 
     seenBackendIds.add(axNode.backendDOMNodeId);
@@ -1243,7 +1259,7 @@ async function handleVomObservation(
     );
     const axNodes = result.nodes ?? [];
     const captured = await captureForVom(deps.cdp, target.tabId, conditionalSurfaceProbe);
-    const scene = buildVomScene(axNodes, captured);
+    const scene = buildVomScene(axNodes, captured, { pageUrl: target.url });
     const rendered = renderVom(scene, {
       maxDepth: params.max_depth,
       maxTokens: params.max_tokens,