Deepfake Detection for Openclaw

A professional-grade deepfake and face-swap detection utility that analyzes media files for AI manipulation using the Scam.ai API.

yuchennnnnnn
v1.0.3
Apr 1, 2026
0
927
0

Install & Download

1. ClawHub CLI

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

npx clawhub@latest install deepfake-detection

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 deepfake-detection 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 Deepfake Detection?

The Deepfake Detection skill is a specialized tool designed for Openclaw Skills users to verify the authenticity of visual media. By leveraging the advanced Scam.ai API, this skill identifies face-swaps and AI-generated faces in both static images and dynamic video content. It provides a secure, streamlined interface for developers and security researchers to integrate deepfake verification into their local workflows.

Built with security in mind, the skill manages API authentication locally and supports both service-specific and universal keys. It removes the complexity of manual API calls, allowing users to simply point the agent to a file path and receive a high-confidence verdict. This is an essential utility for anyone combating misinformation or verifying digital identities in the age of generative AI.

Deepfake Detection Use Cases

  • Verifying the authenticity of social media videos or profile pictures.
  • Identifying face-swapped content in security investigations.
  • Validating user-uploaded media in KYC (Know Your Customer) workflows.
  • Testing the efficacy of AI-generated content against detection algorithms.
  • Quickly screening suspicious media files during digital forensic analysis.

How Deepfake Detection Works

  1. The skill checks for a local Scam.ai API key in the user's home directory.
  2. If no key is found, it guides the user through the Scam.ai registration and key generation process.
  3. The user provides a path to a local image or video file.
  4. The tool performs a POST request to the Scam.ai defense endpoint using curl.
  5. The skill parses the JSON response to calculate a confidence score.
  6. A final verdict (Real, Deepfake, or Suspicious) is presented to the user with the specific confidence percentage.

Deepfake Detection Setup

To start using this skill within Openclaw Skills, ensure you have a Scam.ai API key. You can initialize the setup by running the skill:

# The skill will check for these files automatically:
# ~/.scamai_deepfake_key
# ~/.scamai_universal_key

# To manually save your key with correct permissions:
echo -n "YOUR_API_KEY" > ~/.scamai_deepfake_key && chmod 600 ~/.scamai_deepfake_key

Deepfake Detection Data Schema & Taxonomy

The Deepfake Detection skill manages local configuration and processes external media formats as follows:

Data Type File Extension / Path Purpose
API Key ~/.scamai_deepfake_key Stores the service-specific detection key.
Universal Key ~/.scamai_universal_key Stores the multi-service Scam.ai key.
Image Formats .jpg, .jpeg, .png, .webp, .gif, .bmp Supported static media for analysis.
Video Formats .mp4, .mov, .avi, .mkv, .webm Supported motion media for analysis.
Output JSON Response Parsed for Verdict and Confidence % mapping.

Deepfake Detection Advanced Features

  • Multi-key fallback: Automatically detects and uses universal keys if specific detection keys are missing.
  • Secure credential management: Automatically applies 600 permissions to key files to ensure they are only readable by the owner.
  • Automated path expansion: Handles tilde (~) expansion for easier file referencing in the CLI.
  • Intelligent error handling: Provides specific troubleshooting steps for 401 (Unauthorized) and 429 (Rate Limit) status codes.
  • Session persistence: Saves keys locally so users only need to authenticate once across multiple Openclaw Skills sessions.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*