patens by Patens

Stop losing hours to context switching. Bridge the gap between Chrome and your IDE.

patens is a local memory layer that sends web context—like StackOverflow answers, API documentation, and GitHub issues—directly to your local AI database. It eliminates messy copy-pasting, bloated chat prompts, and AI hallucinations by giving your IDE (Cursor, Copilot, Cline) direct access to your web research via the Model Context Protocol (MCP).

The Old Way (Why you need this):
You're stuck on a bug. You find a great answer on StackOverflow, copy it, switch to your IDE, and paste it into the AI prompt. The AI's response is generic. You switch back to Chrome, copy the official API docs, and paste them again. Your context window is blown, you’ve lost your train of thought, and tomorrow, you’ll have to Google it all over again.

The patens Workflow:
You find the answer in Chrome. You highlight it and press Ctrl+Shift+C.
You switch to your IDE and type: "Fix this using @00_Context_Index.md".
Your AI instantly reads your physical context files, processes your custom research natively, and generates perfect code.

Key Features:

⚡ Zero-Friction Capture: Highlight text on any webpage, press the shortcut, and instantly save it to your local database. No tab switching required.

🧠 Dynamic IDE Mounting: Your IDE dynamically targets your active project. Just tag your master 00_Context_Index.md map to give your AI total recall of your research.

✨ Smart Processing: patens automatically handles deduplication and title resolution behind the scenes.

🔒 100% Local & Private: Powered by a local Python server (SQLite-vec). Your browsing history and API research never leave your machine. No telemetry. No cloud.

How to get started:

Install this extension.

Download the free, open-source patens local server from our GitHub.

Highlight text, hit Ctrl+Shift+C, and send it to your IDE!

(Note: This extension requires the patens local desktop server running on localhost to function properly).