ContextCove — AI Context Curator

ContextCove helps you build the perfect context pack for your AI tools. Browse documentation normally. When you find a useful page, clip it. Discover related links, filter by path or regex, and batch-clip entire doc sections. When you're done, export everything as a single compiled Markdown file or list of URLs — ready to drop into NotebookLM, paste into Claude, or feed to Cursor. AI tools are only as good as the context you give them. But assembling that context — hunting through docs, copy-pasting fragments, cleaning up nav noise — is tedious manual work. ContextCove turns that into a structured workflow. KEY FEATURES: → Clip pages with one click or keyboard shortcut (Cmd+Shift+S) Extracts the actual article content, stripping navigation, sidebars, footers, and boilerplate automatically. → Discover & filter links on any page Find all links on a page, then filter by path prefix, regex, or keyword. Batch-clip the ones you need. → Smart noise removal Strips "On this page" sidebars, duplicate H1 titles, floating code labels, copy buttons, and other documentation cruft that pollutes AI context windows. → Compiled Markdown export Combines all your clips into one clean .md file with a generated Table of Contents. Copy to clipboard or download. Estimated token count shown in real-time. → llms.txt detection Automatically detects when a site publishes an llms.txt file and offers one-click import. → Right-click context menus "Clip this page" or "Clip selected text" from anywhere without opening the side panel. → Works on any site Compatible with any standard documentation portal, API reference page, developer blog, or web-based text guide. HOW TO USE: 1. Click the ContextCove extension icon to open the side panel. 2. Enter your main goal (e.g., "Implement authentication"). 3. Browse any developer documentation page normally and click "Clip Page" (or press Cmd+Shift+S / Ctrl+Shift+S) to capture it. 4. (Optional) Right-click any text selection to clip only the highlighted portion. 5. (Optional) Click "Discover Links" to scan for related sub-pages, apply filters, and batch-clip them. 6. When finished, click "Export" to copy the formatted Markdown or download a text file. WHO IS IT FOR? - Developers building AI-assisted workflows with local LLMs, web-based AI chats, or feed directly into AI-powered code editors. - Prompt engineers assembling reference context for complex tasks - Technical writers packaging documentation for AI review - Anyone who's ever wished they could just drag 20 doc pages into NotebookLM at once