https://k4i.top/tags/tampermonkey/Recent content in Tampermonkey on k4i's blogHugo -- gohugo.ioensky_io@outlook.com (K4i)sky_io@outlook.com (K4i)All content is subject to the license of <a rel="license noopener" href="https://creativecommons.org/licenses/by-nc-sa/4.0/" target="_blank">CC BY-NC-SA 4.0</a> .Sun, 22 Mar 2026 21:49:00 +0800https://k4i.top/posts/mitmproxy-plus-tampermonkey-better-llm-dot-dot-dot-viewer/Sun, 22 Mar 2026 21:49:00 +0800sky_io@outlook.com (K4i)Sun, 26 Apr 2026 16:08:06 +0800https://k4i.top/posts/mitmproxy-plus-tampermonkey-better-llm-dot-dot-dot-viewer/<h2 id="introduction">Introduction</h2> <p>When I debug LLM applications, I often want to see the actual HTTP traffic: the request body, model name, tool calls, streaming chunks, token usage, and the final response. SDK logs are useful, but they are usually either too summarized or too noisy.</p> <p><a href="https://mitmproxy.org/">mitmproxy</a> is great for this because it can intercept HTTP, HTTPS, and WebSocket traffic, then show each flow in <code>mitmweb</code>. The missing piece is presentation: raw JSON is precise, but not pleasant to read when the payload contains long messages or nested tool calls.</p>K4ifeatured imagemitmproxyllmtampermonkeydebuggingproxyworkspace-setup