Show HN: Dropnread – Prepare your codebase for LLMs with a simple drag and drop

I built this tool because I found myself constantly needing to share codebases with LLMs for analysis, and manually copying files was tedious. The tool strips comments to save tokens, and excludes irrelevant files like node_modules, .git and so on. Just drag a project folder, and it combines all files into a single text that fits within your chosen LLM's token limit.

Features:

- Support for GPT-4, Claude, Gemini models with their respective token limits

- Comment stripping for common programming languages

- Configurable file/folder exclusions with sensible defaults

- Token counting

- No installation needed, works in your browser (the code never leaves the browser)

- Respects your privacy

- Open Source MIT License

Try it: https://dropnread.io

Source: https://github.com/yani-/dropnread

Tech stack: Pure JavaScript, Tailwind CSS, no build process needed.

Would love feedback from the HN community!