Legal · plain English

Privacy policy.

Last updated   11 May 2026 Version   0.1.0

Grab Context is a Firefox extension that copies the HTML of a clicked element as Markdown so you can paste it into AI coding assistants. It does not contain analytics, telemetry, error reporting, advertising, or any other code that transmits data off your device.

Data the extension stores

The extension stores one boolean value in chrome.storage.local:

The underlying react-grab library also persists a small UI-preference object (toolbar position, collapsed or expanded, default action) in window.localStorage of each origin you visit. That object stays in the page's own origin and is never transmitted off device.

Data the extension reads

The extension never reads the contents of pages you have not interacted with, and never reads form values, cookies, passwords, headers, or any other page state besides the element you clicked.

Permissions

storage

Required to remember whether the in-page UI is enabled.

activeTab

Required for chrome.tabs.captureVisibleTab so the Screenshot action can capture the current viewport. Invoked only after you explicitly trigger Screenshot.

<all_urls> host permission

Required to inject the selection cursor into whatever page you click on. The injection is purely UI; no page content is read unless you click an element with the cursor armed, and even then the result is only written to your local clipboard.

Third-party code

The extension bundles, at build time, the following MIT-licensed dependencies:

bippy Aiden Bai, transitive
turndown Dom Christie
Phosphor Icons Helena Zhang & Tobias Fried

No remote code is loaded at runtime. All JavaScript is bundled at build time. Anyone can verify this by unpacking the signed .xpi and comparing it against the source repository.

Source & contact

Full source: github.com/Dananz/grab-context. Questions, concerns, or vulnerabilities: please open an issue.