An automated intelligence pipeline that aggregates, scores, and formats tech news from six diverse sources into unified digests for developers and researchers.
The fastest way to install a skill directly from the registry.
npx clawhub@latest install dinstein-tech-news-digest
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
Copy this prompt to OpenClaw to install it automatically.
Help me install dinstein-tech-news-digest using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Get the raw skill files in a ZIP archive.
Tech News Digest is a sophisticated automated intelligence system designed to streamline technical discovery within the Openclaw Skills ecosystem. It employs a unified data source model to collect information from RSS feeds, Twitter/X KOLs, GitHub releases, GitHub Trending, Reddit, and web search engines. This skill solves the problem of information overload by consolidating disparate data streams into a single, high-quality output.
By leveraging this skill, users can transform raw data into actionable intelligence. The system includes a robust quality scoring pipeline and deduplication logic, ensuring that only the most relevant and fresh content reaches your final report. Whether you are tracking LLM breakthroughs or monitoring open-source trends, Tech News Digest provides a structured and professional way to stay ahead of the curve.
To integrate this tool into your Openclaw Skills workflow, follow these steps:
# 1. Install required Python dependencies
pip install -r requirements.txt
# 2. Initialize your workspace configuration
mkdir -p workspace/config
cp config/defaults/sources.json workspace/config/tech-news-digest-sources.json
# 3. Set your API credentials (X/Twitter, Brave, or Tavily are recommended)
export TWITTERAPI_IO_KEY="your_key"
export TAVILY_API_KEY="your_key"
# 4. Execute the pipeline to generate a merged JSON report
python3 scripts/run-pipeline.py --defaults config/defaults --hours 48 --output /tmp/td-merged.json
The skill manages information through a structured metadata taxonomy and local storage:
| Component | Description |
|---|---|
| sources.json | Defines the unified data model for RSS, X/Twitter, and GitHub endpoints. |
| topics.json | Contains labels, emojis, and specific search queries for thematic categorization. |
| archive/ | Stores previous digest titles to ensure perfect deduplication across sessions. |
| merged.json | The primary intermediate data format containing scored articles and metadata. |
| templates/ | Markdown and HTML templates for Discord, Email, and PDF outputs. |
Loading
A focused command-line utility designed for private, offline activity logging and performance analysis.

An automated tool for generating standardized RESTful API documentation from requirements and database schemas.

A comprehensive data-driven intelligence tool for navigating China's 2026 home appliance manufacturing landscape and regional clusters.

An intelligent Chinese mutual fund advisor offering real-time valuation, automated buy/sell suggestions, and advanced portfolio profit tracking.

A comprehensive data-retrieval skill for querying real-time AI model performance, intelligence rankings, and pricing across major industry leaderboards.

A comprehensive intelligent stock monitoring and alerting system featuring seven core technical rules and multi-condition resonance detection.








































