I built this browser extension to solve a daily frustration: copying documentation from websites to feed into AI coding assistants like Cursor, Windsurf, or ChatGPT, only to get a mess of ads, popups, and navigation junk mixed in with the actual content.

LLMFeeder uses Mozilla's Readability.js (same tech as Firefox Reader Mode) to extract just the main article content, converts it to clean markdown with Turndown.js, and copies it to your clipboard with a single keyboard shortcut (Alt+Shift+M or ⌥ ⇧ M). No clicking through popups, no selecting around ads, no fighting with modern web clutter.

# Key features:

• One-key extraction and markdown conversion

• Works on Chrome and Firefox

• Completely local processing (no data sent anywhere)

• Configurable content scope and formatting options

• Open source with clean, documented code

The workflow is simple: hit the shortcut on any documentation page, switch back to your editor, paste clean markdown context, get better LLM responses.

It's solving a real problem I have dozens of times per day, and early users seem to find it useful too. Available on both browser stores and GitHub for manual installation.

# Links:

• Demo: https://www.youtube.com/watch?v=JxHWqszlZDw

• GitHub: https://github.com/jatinkrmalik/LLMFeeder

• Chrome: https://chromewebstore.google.com/detail/llmfeeder/cjjfhhapa...

• Firefox: https://addons.mozilla.org/en-US/firefox/addon/llmfeeder/

Would love feedback from fellow developers who deal with this same workflow friction. What other pain points do you have when feeding content to LLMs?