https://blog.lynxflow.co/en/categories/security/Recent content in Security on Lynx Tech BlogHugo -- gohugo.ioenThu, 10 Sep 2026 09:00:00 +0800https://blog.lynxflow.co/en/posts/wsl-docker-vhdx-shrink-reclaim-disk/Thu, 10 Sep 2026 09:00:00 +0800https://blog.lynxflow.co/en/posts/wsl-docker-vhdx-shrink-reclaim-disk/The numbers didn&rsquo;t add up During a routine disk health check, I hit a discrepancy: Windows showed the C drive at 949 GB with 652 GB used — in the red. But df -h / inside WSL (Ubuntu) reported only 135 GB used. That left 500+ GB unaccounted for. Common sense says deleting a file in Linux should shrink usage on the Windows side too — I&rsquo;d deleted 62 GB of video output the night before, so the C drive should have loosened up considerably. It didn&rsquo;t budge.https://blog.lynxflow.co/en/posts/llm-privacy-self-hosted-gateway-guide/Mon, 31 Aug 2026 00:00:00 +0800https://blog.lynxflow.co/en/posts/llm-privacy-self-hosted-gateway-guide/<img src="https://blog.lynxflow.co/images/llm-privacy-self-hosted-gateway-guide.png" alt="Featured image of post Stop Pasting API Keys and Bank Cards into LLM Relays: A Guide to Self-Hosted Privacy Gateways" />Using a third-party LLM relay to access Claude, GPT, or Gemini is cheap and convenient—but have you considered this: the relay operator can see every single word you send to the model. That includes API keys you casually paste in, bank card numbers, login passwords, ID numbers, medical records&hellip; all sitting in someone&rsquo;s server logs, in plaintext. This isn&rsquo;t paranoia—it&rsquo;s architecture. A relay is fundamentally a reverse proxy: your request hits their server, gets unwrapped