- name
- hfmirror-trending-en
- version
- 1.0.0
- description
- Fetches real-time Hugging Face trending data via the public HF-Mirror API and generates structured Markdown reports in English. Suitable for conversational AI agents.
- author
- shunshiwei
- tags
- license
- MIT
- python
- >=3.6
- dependencies
- []
hfmirror_trending_en (Cross-platform Generic Version)
This Skill enables AI agents to autonomously fetch and parse real-time trending data from HF-Mirror (hf-mirror.com).
Data Source Notice: This Skill calls https://hf-mirror.com/api/trending — a public, login-free REST API provided by HF-Mirror. It does not require any tokens or authorization, nor does it involve any authenticated web scraping or bypassing of access controls.Use Cases
When a user inquires about recent trending models, datasets, or projects on Hugging Face or its mirror. Examples:
- "What are the trending models lately?"
- "What's hot on Hugging Face right now?"
- "Push today's Hugging Face mirror trending list."
- "Help me parse the trending data from HF-Mirror."
Agent Workflow
When processing the above commands, AI agents should follow this standard end-to-end logic:
- Auto-Fetch and Parse: The agent should call the processing script located in the Skill's root directory, utilizing its built-in networking capabilities.
python scripts/summarize.py --fetch [out_path.md]*Note: The script is Python 3 compatible and can be run directly in Windows (PowerShell/CMD), Linux (Shell), or macOS environments.*
- Generate Elegant Reports: The script automatically fetches JSON from
https://hf-mirror.com/api/trendingand generates structured Markdown output in English.
- Smart Delivery: The agent reads the generated file content and presents it as a well-formatted message to the user.
Core Design (Cross-Platform & Environment Decoupled)
- Path Agnostic: Agents can locate
scripts/summarize.pyvia relative paths or Skill environment configurations based on their current context. - Zero Dependencies: The script relies solely on Python 3 standard libraries (
json,urllib,os,sys). It requires no third-party packages, allowing it to run smoothly even in minimal container or CLI environments. - Dynamic Fetch: The built-in
--fetchargument eliminates the need to manually prepare intermediate files, enabling a seamless one-click transition from API to report. - Compliant Access: Uses a named User-Agent (
hfmirror-trending-en-skill/1.0) to identify the request source, adhering to public API best practices.
Core Output Fields Explanation
- Model ID: The unique identifier for the model.
- Downloads & Likes: Metrics reflecting community popularity.
- Parameter Size: Automatically converted (e.g., 7B, 27B) to help users evaluate deployment costs.
- Pipeline Tag: Distinction between different AI domains such as ASR, TTS, OCR, etc.