Browser extensions are often judged by their visible features rather than by the messages their privileged background pages accept. Vimium adds keyboard navigation to Chrome, and its background script accepts requests from content scripts running in frames. Advisory GHSA-vx62-cvr5-x3p4 found that one request handler ignored the frame that sent it.
The affected versions are Vimium 2.4.2 and earlier in the Manifest V3 build. Version 2.4.3 fixes the issue. The advisory rates it moderate because the flaw does not create the first compromise. It turns an existing script-execution problem in a low-privilege frame into script execution in the top-level origin.
Missing frameId changed the target
The background function openUrlInCurrentTab accepts a URL and can call chrome.scripting.executeScript for javascript: URLs. The function supplied a tab ID but omitted frameId. In Manifest V3, that omission defaults to the main frame.
Any frame in which Vimium is active could send the message. A cross-origin iframe with a separate vulnerability could therefore ask the extension to inject code into the parent page. The subframe's weak origin became a bridge to the top-level origin, including its cookies, DOM, and authenticated application state.
The advisory describes a cross-frame scripting and privilege-escalation flaw. It is not a universal remote exploit against every Vimium user. It changes the impact of a separate hypothetical DOM-based XSS in the extension or a hostile frame that can reach the handler.
Extension privilege needs origin-aware routing
Extension messages should identify the sender, the tab, and the frame they are allowed to affect. A background handler should not infer the target from a tab alone when a frame context exists. It should also reject navigation or script schemes that are not explicitly required by the feature.
Upgrade to Vimium 2.4.3. Then inspect extension policy across managed browsers. A user-installed extension may update on its own, while an enterprise-pinned version or an unpacked developer build may remain on an old release. Verify the version from the browser's extension page rather than from the store listing.
Review other extensions that support frames and expose a privileged message API. Look for handlers that accept a URL, a tab ID, or a DOM node without a sender and frame check. The same design mistake can appear in screenshot tools, password managers, translation helpers, and accessibility extensions.
Reduce the impact of the next DOM flaw
Content security policy can limit inline script execution, but an extension's privileged APIs may operate outside the page's normal script boundary. Do not treat CSP as proof that a frame message is safe.
Use managed browser policies to limit extension installation and automatic updates where appropriate. For high-risk administrative browsers, separate extension profiles from production credentials and monitor for unexpected navigation or injected script.
If a browser ran an affected version and a suspicious site was opened in a frame, review the top-level session history, extension permissions, and recent process or file activity on the host. Rotate credentials that were available to the affected browser profile. The fix closes the routing error; the review determines whether the error was used.
Why this is more than a browser nuisance
Keyboard-navigation extensions are often given broad host permissions because they are expected to work on every page. That permission is productive, but it concentrates trust in a background process. A single missing frame identifier can make every tab with a hostile iframe a candidate for cross-origin injection.
The practical control is not to remove all extensions. It is to keep the privileged request path narrow, bind each request to its sender and target, and test the extension against cross-origin frame cases. Browser extension security depends on those small routing decisions as much as on the feature code users see.