llmOverlay

▎ Understand any AI answer without breaking your flow. ▎ ▎ llmOverlay turns long ChatGPT and Claude replies into something you can actually interrogate. Highlight a passage ▎ you don't follow, press the shortcut (Ctrl+Shift+E, or ⌘+Shift+E on Mac), and a side panel opens with a focused ▎ explanation — one that reads the surrounding conversation, not just the few words you selected. Ask follow-up ▎ questions and it keeps the thread. ▎ ▎ Why it's different ▎ Generic "explain this" tools answer in a vacuum. llmOverlay grounds every answer in the actual transcript, so ▎ explanations stay relevant to what you and the model were discussing. ▎ ▎ Features ▎ • Highlight-to-explain on chatgpt.com and claude.ai ▎ • Side-panel Q&A grounded in the real conversation ▎ • Follow-up questions with session memory ▎ • Bring your own API key — Anthropic (Claude) or OpenAI (GPT) ▎ • One-click Clear to wipe the panel history ▎ ▎ Private by design ▎ llmOverlay has no server of its own. Your highlighted text and the chat transcript go directly from your browser to ▎ the provider you chose (Anthropic or OpenAI), using your own API key. Keys and history stay in your browser — ▎ nothing is sent to the developer. Full policy: https://connorbatsell.github.io/LLMOverlay/privacy.html ▎ ▎ Setup ▎ Requires your own Anthropic or OpenAI API key, entered on the extension's options page. Works on Chrome 116+.