Skip to content

Visor

Command: /reversa-visor Phase: Any


🖼️ The forensic illustrator

The forensic illustrator works only with images. Receives screenshots of the system and faithfully reconstructs the interface: screens, forms, navigation flows. Doesn't need the system to be running. Just the photos.


What it does

The forensic illustrator works only with images. You send screenshots and it faithfully reconstructs what's there: screens, forms, navigation flows, interface states. The system doesn't need to be running. Just the photos.

Very useful for systems running in hard-to-access environments: locked production, inaccessible database, legacy server nobody knows how to configure anymore.

Requires image support

Visor only works with models that accept images as input. Claude, GPT-4o, and Gemini support this. Check if your engine supports it before using.


How to use

When you activate Visor, it asks for screenshots:

"[Name], to document the interface, send screenshots of the system screens. You can send one at a time or several at once. Prioritize the main screens and most important flows."

Send the images and it handles the rest.


What it documents per screen

  • Inventory: screen name and purpose, current state (loading, empty, filled, error), usage context
  • Forms: fields, labels, types, placeholders, required/optional, visible validations, action buttons
  • Tables and listings: columns, row actions, pagination, filters
  • Navigation: menus, submenus, breadcrumbs, links
  • Feedback: success/error/alert messages, modals, confirmations, tooltips
  • States: compares the same screen in different states when you provide multiple screenshots of it

What it produces

File Content
_reversa_sdd/ui/inventory.md Complete screen inventory
_reversa_sdd/ui/flow.md Navigation flow in Mermaid
_reversa_sdd/ui/screens/[screen-name].md Detailed spec per screen