AI Agent Hub
Back to skills
📁

Web Search Assistant

Office Efficiency Updated 2026.08.30

Paste the following prompt into your AI chat to install this skill:

Please follow https://skillhub.cn/install/skillhub.md to install @org-02qudk26/web-search.

About this skill

Problem: Obtaining Structured Web Data Programmatically

Automated workflows, data analysis pipelines, and application development often require reliable, scriptable access to public internet information. Manually browsing is inefficient and difficult to integrate. This skill provides a command-line interface wrapping the DuckDuckGo Search API, enabling web searches, news lookups, and image/video retrieval to be executed programmatically with results returned in structured formats.

Core Capabilities

The skill's core is a command-line tool that uses various search types and parameter combinations to meet diverse information needs:

  • Basic Web Search (--type web): Returns a list of web pages with titles, URLs, and descriptions, serving as the most general-purpose method.
  • News Search (--type news): Focuses on recent articles, returning publication dates, source outlets, and summaries, ideal for tracking events or researching current affairs.
  • Multimedia Resource Search:
    • Images (--type images): Allows filtering by size, color, type (e.g., photo, clipart), and layout.
    • Videos (--type videos): Supports filtering by duration and resolution, returning metadata like publisher and duration.

Key Filtering and Control Parameters

  • Freshness Control: The --time-range parameter (e.g., d for past day, m for past month) retrieves recent information, filtering out outdated content.
  • Geographic Targeting: The --region parameter (e.g., us-en) fetches region-specific results.
  • Safe Search: The --safe-search parameter (on, moderate, off) controls the safety level of results.
  • Output Formats: Supports --format text (readable), markdown (documentation-friendly), and json (easy to parse). Results can be saved directly to a file via --output.

Key Workflow and Applicable Boundaries

Typical Usage Flow

  1. Identify Search Intent: Determine if web, news, images, or video are needed, and specify requirements for recency, quantity, and region.
  2. Construct the Command: Combine parameters based on intent. Example: Search for German news about a product from the past week: --type news --time-range w --region de-de "product name".
  3. Execute and Process: Run the command to get results. For subsequent programmatic processing, prefer JSON format output.

Applicable Boundaries and Considerations

  • Data Source Dependency: Results rely entirely on DuckDuckGo's index and ranking algorithms, which may differ significantly in coverage and ordering from other search engines (e.g., Google).
  • Feature Limitations: Does not support advanced search operators like site: or filetype:. Image and video results are typically fewer than web results.
  • Performance Considerations: Avoid making numerous requests in rapid succession to prevent triggering rate limits. Result relevance scoring and ranking are not controllable.
  • Not a Replacement: Unsuitable for scenarios requiring deep customization of crawlers or accessing non-public web pages; specialized web scraping frameworks should be considered instead.

This skill offers a lightweight, direct layer for accessing web data, particularly useful for quickly integrating network information retrieval capabilities into existing toolchains. However, its characteristics and limitations regarding the data source must be understood prior to use.

Use Cases

  • A market analyst needs to periodically collect competitors' latest news reports, user reviews, and product announcements from the web, filtering by time range and saving the structured results as JSON for subsequent analysis.
  • A technical documentation author, while writing an article on a new programming language, needs to gather the latest tutorials, official announcements, and image examples of open-source projects from the web as references.
  • A social science researcher tracking a public issue needs to simultaneously retrieve news reports from both Chinese and English regions (e.g., us-en, uk-en) from the past week to compare descriptions across different sources.
  • A fact-checker verifying a social media claim needs to quickly search for relevant web pages and authoritative news reports, sorting by publication date to view the original source and subsequent developments of the event.

Best For

  • A market analyst who needs to regularly monitor competitors and collect industry news to produce analysis reports.
  • An educator or content creator preparing course materials or a technical blog, needing to gather the latest online references (articles, images, videos).
  • A fact-checker or researcher responsible for information collection and verification, needing to quickly verify information authenticity from multi-region, multilingual sources.
  • A developer or data engineer integrating web search functionality into automated scripts or applications, requiring search results in specific formats (like JSON).