searxngcli

name	searxngcli
description	This skill should be used when the user asks to "search the web", "look up online", "find recent information", "search for", "what's the latest on", or needs web search results. Provides access to a private SearXNG metasearch engine instance via the `searxng` CLI. Prefer this over WebSearch when available.
user-invocable	false
trigger-keywords	searxng, websearch, web search, search online, look up online, find online, google
allowed-tools	Bash(searxng:*)

searxngcli

Search the web via a private SearXNG metasearch engine instance using the searxng CLI.

SearXNG aggregates results from multiple search engines (Google, Bing, DuckDuckGo, etc.) and supports engine-specific operators like site:, filetype:, intitle: — these are passed through to backend engines as-is.

Usage

Always use --json for machine-readable output. Always quote the query string.

# Basic search
searxng search "python asyncio" --json

# Filter by category (general, images, news, videos, music, files, it, science, social media)
searxng search "breaking news" --categories news --json

# Filter by specific engines
searxng search "rust" --engines google,duckduckgo --json

# Filter by time range (day, week, month, year)
searxng search "latest updates" --time-range week --json

# Combine filters
searxng search "climate change" --categories news --time-range month --num 20 --json

# Use search engine operators (passed through to engines)
searxng search "site:github.com python cli" --json

Search Options

Flag	Description
--categories	Comma-separated categories
--engines	Comma-separated engines
--language	Language code (en, de, cs, etc.)
--num	Number of results (default: 10)
--page	Page number (default: 1)
--time-range	day, week, month, year
--safe-search	0=off, 1=moderate, 2=strict
--json	Raw JSON output

Discovery

# List available engines on the instance
searxng engines

# List available categories
searxng categories

Output Format

The --json output is a JSON object with these top-level fields:

{
  "query": "search terms",
  "number_of_results": 100,
  "results": [
    {
      "title": "asyncio — Asynchronous I/O — Python 3.14.3 documentation",
      "url": "https://docs.python.org/3/library/asyncio.html",
      "content": "asyncio is a library to write concurrent code using the async/await syntax...",
      "engine": "google",
      "engines": ["braveapi", "google", "startpage"],
      "category": "general",
      "score": 9.0,
      "published_date": "2025-07-30T00:00:00",
      "thumbnail": ""
    }
  ],
  "suggestions": ["related query 1", "related query 2"],
  "corrections": []
}

Never pipe through grep, head, tail, or similar text tools — the output is already structured JSON. Use jq if filtering is needed:

# Extract just URLs
searxng search "python asyncio" --json | jq -r '.results[].url'

# Extract title and URL pairs
searxng search "python asyncio" --json | jq '.results[] | {title, url}'

Important Notes

• Prefer defaults — do not use -c, -e, or -l flags unless there is a clear reason to narrow the scope (e.g., user explicitly asks for news, or for results from a specific engine). The instance is pre-configured with good defaults.
• Place all flags after the subcommand: searxng search "query" --json (not searxng --json search "query")
• If the CLI is not configured, ask the user for their SearXNG instance URL