ddgs-search

What This Skill Does

The ddgs-search skill provides a high-performance, cost-free gateway to the world's most popular search engines and academic repositories. Unlike standard search integrations that require costly API keys or complex setups, this skill leverages the DuckDuckGo engine library to offer a clean, programmatic way to retrieve live data from Google, Bing, Brave, Yandex, Yahoo, Wikipedia, and Mojeek. Beyond general web searching, it features an specialized arXiv integration for academic and AI/ML research. It is designed as a drop-in, lightweight replacement for web-search-plus, outputting standardized JSON that allows it to serve as a universal backend for other AI agent skills requiring real-time information retrieval.

Installation

Installation is streamlined for rapid deployment across all major desktop operating systems (macOS, Linux, and Windows). First, ensure you have Python 3 installed. You can trigger the automated setup via python3 scripts/install.py, which will automatically fetch the ddgs package, verify CLI availability, and run a connectivity test. If you prefer a manual approach, simply run pip install ddgs in your terminal. For integration with existing OpenClaw workflows, ensure the WEB_SEARCH_PLUS_PATH environment variable is exported to point to your local installation, allowing your agent to route all search requests through this fast, reliable engine.

Use Cases

• Academic Research: Instantly discover the latest peer-reviewed papers using the arXiv integration, complete with metadata like authors and abstracts.
• Agentic Knowledge Retrieval: Provide your AI agent with the ability to verify facts, fetch breaking news, or look up documentation in real-time.
• Market Analysis: Compare results across different search engines (e.g., comparing Google results against Brave or Yandex) to avoid search bias.
• Developer Utility: Easily pipe search results into JSON-processing tools like jq for automated data harvesting.

Example Prompts

• "Search for the latest research papers on 3D gaussian splatting from the last month using arXiv and summarize the findings."
• "What are the current top headlines on Google regarding climate policy in Europe?"
• "Find a comprehensive comparison of Python web frameworks using Brave search and list the pros and cons for each."

Tips & Limitations

To avoid potential issues with interactive shells, always prefer the provided ddgs-search wrapper rather than the raw ddgs CLI, as the latter may trigger pagination prompts that interrupt automated agent workflows. The tool is lightweight and does not utilize heavy browser automation or headless Chrome, making it significantly faster (1-3 seconds per request) than traditional web-scraping methods. While powerful, ensure you verify that your network environment allows outgoing connections to search API endpoints, as some restrictive firewalls may block specific search provider domains.