Magic Mirror for Openclaw

A brutally honest AI agent skill that scans your social media profiles to reveal the gap between your public persona and your actual digital behavior.

sophie-xin9
v1.0.0
Apr 7, 2026
0
646
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install magic-mirror

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install magic-mirror using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Magic Mirror?

Magic Mirror is a sophisticated social media diagnostic tool designed for the Openclaw Skills ecosystem. Unlike generic personality tests, it leverages browser automation to perform a deep-scan of a user's logged-in social media accounts—including Douyin, Xiaohongshu, Weibo, Douban, and Bilibili. The skill focuses on identifying the theatrical gap between a user's Side A (publicly posted persona) and Side B (private likes, bookmarks, and actual consumption habits) to produce a witty and insightful Mirror Report.

By analyzing contradictions—such as claiming a minimalist lifestyle while having a shopping cart full of items—Magic Mirror provides a data-backed reflection of the user's digital footprint. It is built to be entertaining, shareable, and strictly privacy-focused, operating entirely within the user's local environment through the ManoBrowser protocol.

Magic Mirror Use Cases

  • Discovering personal digital consumption patterns and behavioral contradictions.
  • Creating entertaining and shareable social media content based on AI roasts.
  • Auditing self-improvement goals (e.g., books marked as to-read vs. books actually finished).
  • Analyzing cross-platform persona shifts, such as professional behavior on Weibo versus entertainment focus on Bilibili.

How Magic Mirror Works

  1. Dependency Check: The skill automatically verifies the presence of the ManoBrowser skill and its corresponding MCP server configuration.
  2. Platform Confirmation: The agent asks the user to identify which social media platforms they currently have logged-in sessions for in their Chrome browser.
  3. Deep Data Collection: Using platform-specific sub-modules, the agent navigates to profile pages to scrape public posts and private interaction history.
  4. Contradiction Discovery: The AI logic cross-references the collected data points to find dramatic differences between stated interests and actual engagement.
  5. Report Synthesis: A structured Markdown report is generated, featuring data-backed roasts, specific numbers, and shareable social tags.

Magic Mirror Setup

To use this skill within the Openclaw Skills framework, ensure you have ManoBrowser installed and the Chrome extension active. If missing, the skill can auto-clone the dependency:

git clone https://github.com/ClawCap/ManoBrowser.git ./manobrowser

Ensure your .mcp.json or config/mcporter.json includes the chrome_navigate tool. You must be logged into the target social media platforms (Weibo, Douban, etc.) in your active Chrome profile for the automation to access private bookmarks and likes.

Magic Mirror Data Schema & Taxonomy

Magic Mirror organizes its findings in a dedicated local directory structure to ensure privacy and persistence:

File Path Description
mirror-reports/{date}_mirror_report.md The final formatted report containing roasts and data tables.
mirror-reports/{date}_raw_data.json The raw JSON metadata collected from the browser scan.

Data is categorized by Persona (A-side) and Truth (B-side) for each platform, tracking metrics like post frequency vs. bookmark counts.

Magic Mirror Advanced Features

  • Automated environment diagnostic and dependency injection for ManoBrowser.
  • Multi-platform cross-referencing to detect shifts in persona between different social circles.
  • Four distinct reporting styles: Brutal (default roast), Gentle (soft landing), Data-Centric (statistical), and Hell-Mode (maximum sarcasm).
  • Local-first privacy architecture where no social media data is transmitted to external servers.
  • Support for custom JavaScript injection via platform-specific sub-modules for high-fidelity data extraction.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*