Howsembly Capture

Record browser workflows and turn them into Howsembly step-by-step guides.

Turn a browser workflow into a polished process guide without writing every step by hand. Howsembly Capture records the actions you choose during an active capture session, creates step text and viewport screenshots, and sends the result to your selected Howsembly workspace as a private draft. Review captured steps in the side panel, remove unwanted steps, then complete the capture and refine the guide in Howsembly. Features include: - Automatic click and input-step capture - Manual viewport capture and optional meaningful-scroll capture - Optional narrated steps converted to transcript text - Default-on redaction for recognized sensitive fields before upload - Editable highlights, colors, blur regions, sizing, positioning, and cropping - Pause, resume, remove-step, and discard controls - Secure Howsembly sign-in and company selection Capture occurs only after you press Start capture. Narration is off by default, microphone audio is not stored by Howsembly, and captured data is sent only to the Howsembly workspace you select. ## Version 0.3.0 - Added manual viewport capture and optional meaningful-scroll detection. - Added optional narrated steps with visible microphone consent and status. - Added reversible screenshot highlights editable in Howsembly. - Added default-on pre-upload redaction for recognized sensitive form fields and manual blur regions. - Added persistent image cropping and a compact focused capture interface. - Improved editor opening with an exact guide-readiness check and one automatic retry. ## Privacy disclosure Howsembly Capture collects workflow step text, the URLs and titles of pages captured by the user, viewport screenshots, and optional speech transcript text. This data is sent only to the user-selected Howsembly company to create a draft guide. It is not sold or used for advertising. Narrated steps are off by default. When enabled, the extension requests microphone access on a visible permission page. Chrome's speech recognition service may process microphone audio to produce text. Microphone audio is transient and is not stored or uploaded to Howsembly; only resulting transcript text is retained with captured steps. Screenshots are captured only during an active session. Sensitive-field redaction is enabled by default for recognized passwords, PINs, payment-security fields, tokens, and API keys. Those pixels are permanently obscured before the screenshot source leaves the browser. Source screenshots and rendered previews are otherwise uploaded to support reversible highlight and manual blur editing. Users can disable sensitive-field redaction, pause, discard, remove individual steps, disable scroll detection, or disable narration at any time.