771
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill bear2u/my-skills --skill web-search- SKILL.md3.0 KB
Overview
This skill provides DuckDuckGo-based web searches for text, news, and images, with region and time-range filtering. It is useful when built-in WebSearch is limited or when you need news/image results, JSON output, or refined regional/time filtering. The skill runs a simple script and returns structured JSON that can be displayed or processed programmatically.
How this skill works
The skill inspects the user request to determine search type: text (default), news, or images. It runs a command-line script that queries DuckDuckGo with parameters for query, result count, region, SafeSearch, and time period, then returns parsed JSON results. Errors like rate limits or network timeouts are handled with retries or guidance to adjust parameters or install required packages.
When to use it
- Built-in WebSearch is unavailable or restricted in your region
- You need dedicated news search or recent news within a specific time range
- You need image URLs or image-focused results
- You require programmatic JSON output for automation or saving results
- You need region-specific results (e.g., Korea, Japan, UK)
Best practices
- Specify search type explicitly (text, news, images) to get targeted results
- Use region codes (kr-kr, us-en, jp-jp, uk-en) for localized relevance
- Limit result count (-n) to avoid rate limits and speed up responses
- Use time period flags (d, w, m, y) for recent or historical filters
- Include search operators (site:, filetype:, "exact phrase") for precision
Example use cases
- Quickly gather recent Korean news about a policy using region=kr-kr and period=w
- Collect image URLs for design inspiration with an images search and n=10
- Save top 20 research hits as JSON for further processing or citation
- Search a specific site for PDFs using site:example.com filetype:pdf
- Get rapid fact-checking links when built-in search is blocked in your area
FAQ
Only the query is required. Other parameters like type, count, region, SafeSearch, and period are optional with sensible defaults.
How do I avoid rate limits?
Reduce the number of results (-n), add short delays between queries, or retry after a pause if you hit a rate limit.
What if the script reports missing packages?
The script attempts to auto-install. If it fails, run pip install -U ddgs manually and retry.