magic-mirror-demo

Capture camera photos, score portraits, and generate queen videos via ArtsAPI.

41|2|Updated Feb 17, 2026
One-click install
npx skills add https://github.com/wanhaoniu/MomoAgent --skill magic-mirror-demo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: magic-mirror-demo
Source: https://github.com/wanhaoniu/MomoAgent/tree/main/skills/magic-mirror-demo
Command: npx skills add https://github.com/wanhaoniu/MomoAgent --skill magic-mirror-demo

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires opencv-python, numpy, and includes scripts (resource) components.

What problem does it solve?

This skill streamlines a local portrait workflow by capturing photos with a camera, selecting the best candidate from the workspace, and generating a queen transformation video via ArtsAPI, all within a self-contained demo.

Core Features & Use Cases

  • Local camera capture via the soarmmoce-real-con integration and a deterministic snap script, saving images into the skill's workspace.
  • Quality-based candidate scoring using face detection, sharpness, brightness, and composition to select the most portrait-worthy photo.
  • Queen-video generation via ArtsAPI, with an option to upload the source image to a public URL and save the resulting video locally.

Quick Start

Run the magic-mirror-demo to capture a photo, pick the best candidate, and generate the queen video.

Frequently Asked Questions about magic-mirror-demo

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

FAQPage Schema
How do I capture and score portrait photos using a local camera?

To capture and score portrait photos, run the magic-mirror-demo to trigger the local camera, save images to the workspace, and automatically evaluate face detection, sharpness, brightness, and composition to select the best candidate.

How does face detection and image scoring work for selecting the best photo?

Image scoring evaluates captured portraits using face detection, sharpness, brightness, and composition metrics, analyzing each photo in the workspace to programmatically identify and select the most portrait-worthy candidate for video generation.

Do I need Python3 and ArtsAPI CLI to generate queen videos?

Yes, generating queen videos requires Python3, the ArtsAPI CLI, the soarmmoce-real-con camera integration, and a designated workspace to store captured photos and output the final generated video files locally.

What's the best way to automatically pick the best portrait from multiple camera shots?

The best way to pick a portrait is using the built-in candidate scoring system, which ranks multiple local camera shots by face detection accuracy, sharpness, brightness, and composition to isolate the highest quality image automatically.

Can I upload the selected portrait to a public URL for ArtsAPI video generation?

Yes, the workflow includes an option to upload your selected source image to a public URL, which is then used by ArtsAPI to generate the queen transformation video before saving the final output locally.

Why does my local camera portrait capture fail to select a usable photo?

Portrait capture may fail to select a photo if the local camera images score too low in face detection, sharpness, brightness, or composition, preventing the system from isolating a portrait-worthy candidate for video generation.