ClickToDoc

ClickToDoc records what you do in your browser and turns it into a polished, share-ready guide. Click around like normal; every step lands with a screenshot, a caption, and a marker showing exactly where you clicked. Then edit, enhance, and export - HTML, Word, PDF, Markdown, an animated GIF, or a JSON any AI agent can read. Made for people who write the same how-to twice a week. Enhance with YOUR AI (bring your own key - or no AI at all) Connect the AI you already use - OpenAI, Google Gemini, Anthropic Claude, OpenRouter, or a fully local model via Ollama / LM Studio - and one click can polish every title, write the descriptions, translate the whole guide, annotate the screenshots by sight, redact sensitive data, or merge near-duplicate steps. Your key stays on your machine, requests go straight from your browser to your provider, screenshots are sent with your redactions already burned in, and every change lands in a review diff you approve (one Ctrl+Z undoes it all). No ClickToDoc servers involved - because there are none. Capture, intelligently Combine many clicks into one step: Multi-click mode collects a whole dropdown-and-menu sequence into a single screenshot with numbered markers - or let your AI merge duplicate steps after the fact. Narrate as you go: turn on Voice Notes and what you say lands on the step you're on, live, via the browser's own speech engine. Typing, tab switches, text selections and dropdown choices are captured as proper steps; sensitive fields are masked; smart-redact flags credit cards, emails and phone numbers. A real editor, not a form Crop, annotate (arrows - straight, curved or hand-drawn - highlights, boxes, ellipses, freehand, text bubbles with movable tails, numbered markers, blur, focus dim, zoom callouts, smart erase that blends into the background), rearrange steps by drag and drop (list or grid view), insert headings and notes anywhere, duplicate steps, paste images from your clipboard as steps, replace any screenshot, and move or copy steps between workflows. Everything is customisable: colours, stroke widths, text sizes, click-marker styles, spotlight and smart-zoom per step - and what you see in the editor is pixel-identical to every export. Export anywhere Self-contained HTML (one file, screenshots embedded), Word DOCX (cover page, table of contents, your own template or branding), print-ready PDF, Markdown for wikis and Notion, an animated GIF walkthrough, all step images as a ZIP, or workflow JSON that any AI agent can read and edit (the format is openly documented). Optional publish-to-your-own-server and Google Drive upload for self-hosters. Local-first, provably No account. No backend. No telemetry. Your recordings live in your browser's storage; optional scheduled JSON backups go to your own Downloads folder and restore on any machine. The optional AI and Drive integrations talk only to servers YOU configure with YOUR credentials. What this is not Not a screen recorder: crisp screenshots per click, not video (the GIF .mp4 export animates them if you want motion). Not a cloud service: no website login, no shared library, no seat pricing. Not another AI subscription: bring your own key, or use none at all - every feature works without AI.