You already know how to do the thing. ArkScribe watches you do it once, then turns the clicks, the typing and the scrolling into a finished guide. Recorded in your browser, exported to PDF, HTML, Markdown or a narrated video. Nothing is uploaded.
Most documentation is written twice. Once badly, by hand, with the screenshots in the wrong order. Then again next quarter, when the feature changed and nobody updated it. ArkScribe removes the first pass entirely. You perform the task once, as you normally would, and it captures every action with the context needed to explain it. You clean up the timeline, export it in whatever format the reader needs, and the guide is done. When the feature changes, you record it again. That is the whole workflow.
Onboarding flows, feature walkthroughs, release notes. Recorded on the actual build, not described from memory. Works across desktop apps, web apps, Android and iOS.
You have already solved the problem to answer the ticket. Record the fix while it is fresh, and the help article writes itself. No second pass, no screenshot hunting.
Reproduction steps that are accurate by construction, because they are the steps you actually took. Attach them to the ticket and the next person can follow them exactly.
Hand a colleague the exact path to reproduce a bug, including the timing, instead of a paragraph describing it.
Turn any procedure into a guide staff can follow. Built for people who learn by watching, not by reading a PDF.
Course material built from the actual interface, so a student following the guide sees the same screen you recorded.
Feature demos and pitch material recorded from the shipping product, so the deck and the software never disagree with each other.
Product footage for launch posts and sales pages, produced by the team that built the thing rather than requested from elsewhere.
Handoff documentation that makes you look like a bigger operation. SOPs and walkthroughs exported on your own domain.
Process documentation you can hand to a client as evidence of how you work, exported on your own domain.
Recognises what you clicked, whether it was a button, a link or a form control, and names it for you.
A full-page shot on every action, with a cursor marker so the reader knows where to look.
Typed text is recorded and consolidated into one clean entry per field, not one line per keystroke.
Notices when something scrolls into view and captures it, so lazy-loaded steps still get documented.
Finds field names from HTML label attributes, ARIA roles and nearby text.
Choose JPEG, PNG or WebP, and set quality and dimensions, before you export.
Drag to reorder, delete the noise, or merge several actions into one step. Everything renumbers itself.
Click any step title or description and rewrite it in place. Auto-saved as you type.
Arrows, text, shapes, cropping and blur, for the detail a screenshot alone will not carry.
Saved in your browser as you work. A closed tab does not lose the guide.
PDF, two HTML variants, two Markdown variants and a PNG folder, packaged in a single ZIP.
Convert the guide into a 2D or 3D presentation video with text-to-speech narration.
Upload your own HTML template and the export carries your branding instead of ours. Pro.
Save and load projects from a repository, and publish to GitHub Pages in one click.
No account, no login, no cloud upload. Screenshots and guides are processed and stored in your browser, and never transmitted. The permissions the extension asks for exist to make that possible: local storage to save your work, activeTab to read the page you are recording, unlimitedStorage so a long guide with high-resolution screenshots fits, and notifications to tell you when an export finishes. There is no analytics, no usage tracking, and no telemetry. If you clear your browser data, your guides go with it, and we cannot recover them because we never had them.
Unlimited steps, no watermark, and your own HTML layouts so the export is branded yours. You choose the amount.
Is ArkScribe really 100% private?
Yes. Screenshots, descriptions and guides are processed and stored locally in your browser. Nothing is uploaded to any server, there is no analytics, and no account exists to tie data to.
What is the difference between Free and Pro?
Free gives you unlimited recordings capped at ten steps per guide, every export format, the full editor and ArkPres video generation, all watermarked. Pro removes the watermark and the step cap and adds custom HTML layouts for complete white-label output. It is a one-time payment, not a subscription.
Which browsers are supported?
Chromium-based browsers: Chrome 88 and above, Edge, Brave and Opera. Firefox and Safari are not supported, because their extension APIs work differently.
Can I document mobile apps?
Yes. It records whatever the browser can display, so desktop applications, web apps, Android apps and iOS apps all work. Emulators and responsive views included.
What does an export actually contain?
A single ZIP holding a print-ready PDF, HTML with embedded images, HTML with separate images, Markdown with embedded images, Markdown with separate images, and a folder of every screenshot as a PNG.
Does it use AI to write the descriptions?
No. Descriptions come from detection: the element you clicked, the label associated with the field, the action that happened. Nothing is invented, which is why the output stays accurate on internal tools and unreleased features.
Why do I need a Pro license if it is a browser extension?
Validation is a single one-time check when you activate. Your guide content and screenshots are never uploaded, only the licence key is confirmed.
Record the task once and you have a guide, a video and a set of release notes. No account, no upload, no watermark on the free tier. Add it to Chrome and try it on something you already documented badly.