Fetch ranked web search results from a self-hosted SearXNG endpoint.
npx skills add https://github.com/Emrebener/Sublime-Skills --skill ss-web-search
Please help me install this Agent Skill. Skill: ss-web-search Source: https://github.com/Emrebener/Sublime-Skills/tree/main/skills/web-utilities/ss-web-search Command: npx skills add https://github.com/Emrebener/Sublime-Skills --skill ss-web-search
ss-web-search removes the need for agents to rely on external search services or an MCP search server by returning ranked web results from your own SearXNG instance.
Configure your SearXNG endpoint via the SEARXNG_URL environment variable, then ask your agent to run: ss-web-search "openai pricing" --count 5 --category general --time month.
High-intent search queries and answers about installing and using this skill.
You can fetch ranked web search results by querying a self-hosted SearXNG JSON endpoint. This approach removes the need for an MCP search server, returning titles, URLs, and snippets directly to your agent workflow.
Yes, SearXNG search results can be filtered by category such as general, news, images, or videos. You can also apply recency constraints like day, week, month, or year, and select specific languages or regions.
Yes, fetching JSON output from SearXNG requires Node.js 18 or higher with fetch support. The tool reads the SEARXNG_URL environment variable or a local config.json to locate your self-hosted endpoint.
You can return JSON arrays from a SearXNG query by specifying JSON output format in your request. This produces programmatic data for downstream use instead of the default human-readable ranked text list.
The best way to use SearXNG for documentation lookup is querying a self-hosted JSON endpoint with category and time filters. This provides up-to-date web sources with ranked titles and snippets for your agents.
vikrant-projectAggregate search results from over 70 engines via SearXNG.
guyruSearch Bing web results without an API key using Python 3 stdlib.
kryptobaseddevGather and synthesize research from web searches, Context7 docs, and codebases into a citation-backed brief.
Kirankumar2604Access external web APIs through Replit-managed proxying and automatic authorization.
