AI Code Prep - Electron Version
AI Code Prep: Speed up vibe coding with AI chats (Much less Copy & Paste!)
This is a Windows Installer file for it. Install, open it, and when you need to provide an AI chat with your code/context, just drag and drop your project folder onto the window. Saves so much time since you don't have to go to each file and copy the code, paste into chat.
Stop wasting time manually selecting files for your LLM! Copy, paste, copy, paste, I know you hate it. I did, which is why I made this.
Drag and drop a project folder with code files and it smart-selects the likely files you probably want / need to give the AI for context. Then you can fine tune it by taking off some files, adding some. Sometimes if you want something done good, you gotta do it yourself.
Drop a folder on it, adjust, click process (automatically copies it to clipboard), then paste into a LLM chat, or download a .md, .json, or .txt file.
Key Features:
- Smart Auto-Checkboxing: AI Code Prep automatically selects the most relevant files based on a customizable YAML configuration. It prioritizes code files, intelligently excludes irrelevant directories (like node_modules), and lets you fine-tune with include/exclude patterns. This saves you tons of time and effort.
- Right-Sized Context: No more overwhelming your LLM with too much information or starving it with too little. AI Code Prep ensures your AI gets just the right amount of context for optimal performance.
- Interactive File Tree: Want manual control? The interactive file tree lets you easily override the auto-selection, with visual cues for excluded items.
- One-Click Copy/Download: Instantly grab the prepared context for your LLM.
- Modern UI: Dark mode, Material-UI, font selection, and drag-and-drop support.
- It doesn't send your code to remote servers - like a lot of tools. This is local only.
Future features: Compressed context summaries, token count, integrated browser with auto-saving and editing files. Guide/knowledgebase for which AI models are good for which type of task - like frontend, C++, debugging, crowdsourced.