reversa-visor

Documents legacy system interfaces from screenshots by extracting components, layouts, and navigation flows.

Updated May 22, 2026
One-click install
npx skills add https://github.com/GleisonOliveira/dockerpilot-mcp --skill reversa-visor-gleisonoliveira
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: reversa-visor
Source: https://github.com/GleisonOliveira/dockerpilot-mcp/tree/main/.agents/skills/reversa-visor
Command: npx skills add https://github.com/GleisonOliveira/dockerpilot-mcp --skill reversa-visor-gleisonoliveira

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Reverse-engineering a legacy system's UI is hard when the system cannot be run or its source is unavailable. This Skill turns user-provided screenshots into structured interface documentation without requiring the system to be running. ## Core Features & Use Cases - Screen Inventory: Catalogs each screenshot with its purpose, state (loading, empty, filled, error), and usage context. - UI Element Extraction: Documents forms, tables, navigation menus, and feedback elements (modals, messages, tooltips) per screen. - Navigation Flow Mapping: Produces a Mermaid diagram of navigation flows across screens and maps each screen to a spec unit based on configurable granularity (module, endpoint, use-case, feature, hybrid, custom). - Use Case: A team modernizing a legacy application sends screenshots of the main screens; the Skill generates per-unit screens.md specs, a global inventory.md, and a flow.md navigation diagram for the reverse-engineering documentation set. ## Quick Start Send screenshots of the legacy system's main screens and ask the agent to document the interface, mapping each screen to its specification unit.

Frequently Asked Questions about reversa-visor

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I document a legacy system UI without access to the running system?▼

Provide screenshots of the system's screens to the agent, which extracts components, layouts, states, and navigation flows into structured Markdown specs. The system does not need to be running; only images are required.

How are screenshots mapped to specification units?▼

Each screen is mapped to a unit based on the granularity setting in .reversa/config.toml: module, endpoint, use-case, hybrid, feature, or custom folders. Ambiguous mappings are confirmed with the user before saving.

What output files does the screen documentation generate?▼

Per unit it writes the original screenshots and a screens.md spec. Globally it writes ui/inventory.md listing all screens with their units and ui/flow.md containing a Mermaid navigation diagram.

Does this skill work with AI agents that lack image support?▼

No. It requires a model with image support, since the entire analysis is based on reading screenshots. It is compatible with Claude Code, Codex, Cursor, Gemini CLI, and other Agent Skills-compatible agents with vision.

What happens if I send the same screenshot twice?▼

Existing screenshots and specs are never deleted or overwritten. A duplicate screen is saved with a numeric suffix, such as tela-2.png, following the non-destructive directive.