FBFeeder

A browser extension to help manage your Facebook feed.

Installation

Chrome / Edge / Brave

1. Open your browser and navigate to chrome://extensions.
2. Enable Developer mode in the top right corner.
3. Click Load unpacked.
4. Select the directory containing this repository.

Safari (macOS)

To use this extension in Safari, you will need to convert it using the standard xcrun safari-web-extension-converter tool or load it as an unpacked extension if you have the Safari Web Extension development features enabled.

1. Open Safari.
2. Go to Develop > Allow Unsigned Extensions.
3. Go to Settings > Extensions.
4. Check FBFeeder.

Note: You may need to build a wrapper app for permanent installation in Safari.

Firefox

1. Open Firefox and navigate to about:debugging.
2. Click This Firefox.
3. Click Load Temporary Add-on...
4. Select the manifest.json file in this repository.

Development History

Initial thoughts

From the structure of the page, the initial plan is simple

1. find the "div" that holds each post
2. decide if it's an Ad.

While the initial implementation is more heuristic, it kind of works by false-positively identifying some posts as Ad (so be it, FB post is not essential).

Time goes on and it became a game and I kind of want to know what's going on.

Obfuscation

When rendered, it's easy for us to see Sponsored, Follow, Join, etc. However, they're not presented in the DOM as-is. FB does go to some lengths to make it hard.

A few interesting techniques:

SVG Obfuscation

is used and the implementation is not inline. so the simple innerText approach doesn't work.

Flexbox Obfuscation (Dec 2025 finding)

Facebook found a way to render "Sponsored" as "tSsnnrd" (or a variation of it) by:

1. Putting characters in a display: flex container.
2. Mixing in decoy characters (hidden with position: absolute).
3. Randomly shuffling the DOM order of the characters.
4. Using CSS order: N style to visually re-arrange them back to "Sponsored".

The fix involves finding display: flex spans, filtering out absolute/hidden children, sorting the rest by their computed order, and essentially OCR-ing the text back.

Notes

Well, the codes are there for you to see and modify. Exactly the way I want to load any extension to my browser.

It works currently and will likely continue to work (the skeleton of the code remain the same since May 2025) until FB release another major version of UI.