langextract-search

What This Skill Does

LangExtract Search is a powerful automation skill designed for OpenClaw that synthesizes web search, multi-source data retrieval, and AI-driven structured extraction into a single, cohesive workflow. By combining the precision of Zhipu AI's search engine with the breadth of DuckDuckGo’s search capabilities, it enables users to fetch information from the internet and transform raw search results into organized, structured data formats automatically. Whether you are performing market research, collecting competitive intelligence, or gathering real-time data for AI model training, this tool streamlines the research pipeline by handling the complexities of searching, parsing, and data structuring within one execution step.

Installation

To install this skill, use the ClawHub command-line interface. Run the following command in your terminal: clawhub install openclaw/skills/skills/luw2007/langextract-search

Before execution, ensure your environment is prepared by installing the necessary dependencies: pip install requests ddgs langextract

If you are a first-time user, simply running the skill without a baseUrl in openclaw.json will trigger an interactive setup, allowing you to configure your preferred model settings, which will then be saved automatically to conf.json.

Use Cases

• Market Research: Automatically search for competitor pricing or industry news and extract relevant data points into a structured JSON format.
• Trend Tracking: Monitor specific topics across multiple search engines and aggregate findings for daily reports.
• Content Curation: Scrape web content based on complex search queries and filter the information based on date, region, or source credibility.
• Automated Reporting: Compile large search results into summaries using your configured LLM, bypassing the need for manual browsing.

Example Prompts

1. "Search for the latest trends in renewable energy for the past month and extract the top 5 companies mentioned into a JSON file."
2. "Find recent news about AI regulation in the US using DuckDuckGo, extract the core arguments, and save the result as a local data object."
3. "Look up detailed specifications for the latest smartphone releases and extract the display and battery metrics using the LangExtract engine."

Tips & Limitations

• Efficiency: For best results, use the --verbose flag during your first few test runs to ensure that the search query targeting and extraction logic are returning the expected data quality.
• Performance: The skill relies on external APIs (Zhipu/DuckDuckGo). If you experience latency, consider checking your proxy settings in conf.json or adjusting the --ddg-max-results parameter to a lower number.
• Configuration: Advanced users should leverage the references/search-params.md documentation to fine-tune search behavior, especially when setting geographic constraints or time-bound filters for more localized information gathering.