Context Packer — pages & PR threads to LLM-ready Markdown

Turn any page, GitHub PR thread, or code file into clean LLM-ready Markdown with token estimates — copied in one click.

Stop fighting nav bars, sign-in banners, and broken formatting every time you paste web content into ChatGPT, Claude, or Gemini. Context Packer turns the page you're on into a clean, structured Markdown "context pack" copied straight to your clipboard, with a token estimate so you know what it costs before you paste. WHAT IT PACKS ⚡ GitHub pull requests - title, state, description, the full conversation with authors and timestamps, review threads with file context, and the changed-files list. Works on private repos and GitHub Enterprise because it reads the page you're looking at — no tokens, no API access, nothing leaves your browser. ⚡ GitHub code files - the raw file as a single fenced code block, headed by repo, path and branch. No line numbers, no duplicated rendering, no UI noise. ⚡ Articles & docs - clean readable extraction (Readability) with headings and links preserved as Markdown. ⚡ Any selection or full page - when you just want exactly what you see. WHY IT'S DIFFERENT Every pack starts with a header comment (title, URL, capture date) so your LLM(and future you) knows where the context came from. A token estimate shows before you paste. And everything is processed locally in your browser: no accounts, no servers, no data collection, ever. PRIVACY Context Packer runs only when you click it (activeTab). Its only background behavior is opening the welcome tab once on install. It never reads pages you don't invoke it on, makes no network requests, and collects nothing. Your pages and packs never leave your machine except via your own clipboard.