ai-daily-digest

What This Skill Does

The ai-daily-digest skill is a powerful automation tool designed to curate, filter, and synthesize content from the 90 elite technology blogs recommended by Andrej Karpathy. Instead of manually scanning endless RSS feeds, this skill leverages AI models (such as Gemini or DeepSeek) to evaluate article relevance and technical depth. It processes raw data into an insightful, structured Markdown digest featuring Chinese-translated titles, intelligent category grouping, visual trend analysis via Mermaid charts, and a visual tag cloud to highlight recurring themes. The tool is designed to be highly configurable, allowing users to define time windows and the number of articles they want to see, ensuring that information consumption remains efficient and high-signal.

Installation

To install this skill, run the following command in your terminal: clawhub install openclaw/skills/skills/royxiao08/ai-daily-digest

Ensure that you have bun and a valid Gemini API Key (or an OpenAI-compatible alternative) before triggering the command, as these are required for the AI-driven synthesis and content translation tasks.

Use Cases

• Software Engineers: Stay updated on major architectural breakthroughs and niche development trends from top-tier engineering blogs without browsing dozens of websites.
• AI Researchers: Rapidly parse the latest technical developments and papers discussed by top industry voices.
• Tech Enthusiasts: Consolidate a week's worth of reading into a readable, concise summary delivered daily or weekly.
• Language Learning: Use the summary features to practice reading technical concepts by toggling between English and Chinese outputs.

Example Prompts

1. "Run the daily digest for the last 48 hours."
2. "I need a tech news summary for my morning routine, can you start the /digest process?"
3. "Please generate an AI blog digest with the latest 15 articles."

Tips & Limitations

• Pro-tip: The skill supports persistent configuration in ~/.hn-daily-digest/config.json. Once you set your preferred time range and quantity, you can trigger future runs with a single confirmation, making it a true 'set and forget' productivity utility.
• Limitations: The quality of the summary depends on the API provided. While Gemini is the recommended primary, using a high-quality model is necessary for accurate translations and coherent category grouping. Ensure your internet connection is stable, as fetching 90 RSS feeds involves multiple concurrent network requests.