image-fetcher

name	image-fetcher
description	根据关键词搜索并获取相关图片
version	4.0.0
domain	office
layer	content
tags	["slides","image","fetch","search"]
taskTypes	["slides-generation"]
priority	3
author	genesis-ai
source	local
execution-mode	provider

Image Fetcher

This skill is implemented as a NestJS Provider (code-based tool). This SKILL.md provides metadata only — execution is handled by the corresponding .skill.ts file.

Image fetching skill that searches for and retrieves high-quality images based on keywords. Supports Unsplash API (free 50 requests/hour) with automatic fallback to local image library when API is unavailable.

Key Features

• Searches Unsplash for relevant images based on keywords
• Automatically extracts keywords from titles and content
• Falls back to categorized local image library when API is unavailable
• Supports multiple image sizes (small, medium, large)
• Provides orientation options (landscape, portrait, squarish)

Image Categories

Includes fallback images for: business, technology, data, team, growth, innovation, weather, city, shopping, lifestyle, network, and nature.

Input Requirements

• keywords: Array of search keywords
• size (optional): Image size - small (400), medium (800), or large (1600)
• orientation (optional): Image orientation - landscape, portrait, or squarish
• count (optional): Number of images to return

Output Structure

Array of image results, each containing:

• id: Unique image identifier
• url: Image URL with specified size
• thumbnailUrl: Thumbnail URL
• width and height: Image dimensions
• description: Image description (if available)
• author and authorUrl: Attribution information (if available)