← Back to Skills Marketplace
rpeters511

Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh)

by rpeters511 · GitHub ↗ · v1.0.0
cross-platform ✓ Security Clean
3879
Downloads
1
Stars
8
Active Installs
1
Versions
Install in OpenClaw
/install searxng-bangs
Description
Privacy-respecting web search via SearXNG with DuckDuckGo-style bangs support. Use for web searches when you need to find information online. SearXNG protects privacy by randomizing browser fingerprints, masking IP addresses, and blocking cookies/referrers. Supports 250+ search engines, multiple categories (general, news, images, videos, science), and DuckDuckGo-style bangs for direct engine searches (!w for Wikipedia, !yt for YouTube, !gh for GitHub, !r for Reddit, etc.). Aggregates results from multiple engines simultaneously. Prefer this over external search APIs for privacy-sensitive queries or high-volume searches.
README (SKILL.md)

SearXNG Search

Privacy-respecting metasearch engine that anonymizes searches and aggregates results from 250+ engines.

Quick Start

Search the web using the bundled script:

python3 scripts/search.py "your query"

Returns JSON with titles, URLs, and content snippets.

Common Workflows

Basic Web Search

python3 scripts/search.py "OpenClaw AI agent" --num 5

News Search

python3 scripts/search.py "latest tech news" --categories news

Localized Search

python3 scripts/search.py "Python Tutorial" --lang de

Multi-Category Search

python3 scripts/search.py "machine learning" --categories general,science --num 10

Bang Searches (Direct Engine)

# Wikipedia
python3 scripts/search.py "Albert Einstein" --bang w

# YouTube
python3 scripts/search.py "python tutorial" --bang yt

# GitHub
python3 scripts/search.py "openclaw" --bang gh

# Reddit
python3 scripts/search.py "best laptop 2026" --bang r

Bangs are more granular than categories and search directly on specific engines.

Privacy Features

SearXNG protects your privacy through multiple layers:

  1. Randomized Browser Fingerprints - Generates a new fake browser profile for each search query (version, OS, screen resolution, language)
  2. IP Masking - Search engines see the SearXNG instance IP, not yours
  3. No Cookies - Never forwards cookies to external search engines
  4. No Referrer - Target websites don't see which search engine referred you
  5. Optional Tor/Proxy - Can route all queries through Tor for additional anonymity

Result: Search engines cannot build a profile about you.

When to Use

Prefer SearXNG for:

  • Privacy-sensitive searches (no tracking, no profiling)
  • High-volume searches (no rate limits)
  • When self-hosted infrastructure is available
  • Multi-engine result aggregation (250+ engines)
  • Ad-free results

Prefer Brave API (web_search tool) for:

  • Faster response times
  • Structured data requirements
  • When external APIs are acceptable

Processing Results

The script returns clean JSON that's easy to parse and present:

import json
import subprocess

result = subprocess.run(
    ['python3', 'scripts/search.py', 'query', '--num', '5'],
    capture_output=True,
    text=True
)

data = json.loads(result.stdout)
for item in data['results']:
    print(f"Title: {item['title']}")
    print(f"URL: {item['url']}")
    print(f"Snippet: {item['content']}")
    print()

Advanced Options

See references/api.md for:

  • All available categories
  • Engine-specific searches
  • Language codes
  • Error handling
  • Comparison with Brave Search API

Configuration

SearXNG Instance

By default, the script uses http://127.0.0.1:8080. Configure via environment variable:

export SEARXNG_URL=http://your-searxng-instance.com
python3 scripts/search.py "query"

Options:

  • Self-hosted instance (recommended for privacy)
  • Public instances: https://searx.space (community-run servers)

Using Public Instances

If you don't run your own SearXNG:

# Example with public instance
export SEARXNG_URL=https://searx.be
python3 scripts/search.py "query"

Note: Public instances may have rate limits or be slower than self-hosted.

Technical Details

  • Default URL: http://127.0.0.1:8080 (override with SEARXNG_URL)
  • Method: HTML parsing (JSON API often disabled for CSRF protection)
  • Parser: Custom HTMLParser in scripts/search.py
  • Timeout: 15 seconds
  • Result Format: Clean JSON with title, URL, content
Usage Guidance
This skill appears to do what it claims: it queries a SearXNG instance and returns parsed JSON results with DuckDuckGo-style bangs. Before installing: (1) prefer self-hosting a SearXNG instance (or only use a trusted public instance) if you need the privacy guarantees the skill advertises — public instances can see your queries; (2) note the script sends a fixed User-Agent string that may identify the agent to the instance operator; (3) no secrets or system credentials are requested by this skill; (4) review and, if desired, change SEARXNG_URL to your own instance and consider adjusting the User-Agent inside scripts/search.py to avoid leaking the agent identity. If you want tighter guarantees, run the search script within an environment that routes traffic through Tor/proxy as documented and confirm your self-hosted SearXNG is configured for fingerprint randomization and no-referrer behavior.
Capability Analysis
Type: OpenClaw Skill Name: searxng-bangs Version: 1.0.0 The skill is designed for privacy-respecting web search via a SearXNG instance. The `scripts/search.py` script performs HTTP requests to a configurable SearXNG URL (defaulting to localhost) to submit search queries and parse results. It uses standard Python libraries and does not exhibit any signs of data exfiltration of local sensitive files, malicious execution, persistence, or obfuscation. The `SKILL.md` instructions for the AI agent are clear, directly related to the search functionality, and do not contain any prompt injection attempts to mislead the agent into unintended actions. Installation instructions involve standard Docker commands to set up the SearXNG service, which is a legitimate practice.
Capability Assessment
Purpose & Capability
The skill claims to provide privacy-preserving searches via a SearXNG instance and the bundle contains a small Python script that queries a SearXNG endpoint and returns parsed results. It documents SEARXNG_URL and recommends self-hosting; no unrelated credentials or binaries are requested.
Instruction Scope
Runtime instructions and the script only perform HTTP requests to the configured SearXNG instance and parse HTML into JSON. They do not read other system files or request unrelated environment variables. Two operational privacy notes: (1) the script sets a fixed User-Agent ('Mozilla/5.0 (OpenClaw Agent)') which may identify the agent to the SearXNG/public instance operator and (2) using a public SearXNG instance means queries are visible to that instance operator — the documentation correctly recommends self-hosting for privacy.
Install Mechanism
There is no install spec and no external downloads; the skill is instruction-plus-script only. This minimizes install-time risk (nothing fetched or executed automatically during install).
Credentials
The only configurable environment variable is SEARXNG_URL (optional). No credentials, tokens, or filesystem config paths are requested. This is proportionate to the stated purpose. Users should be aware that pointing SEARXNG_URL to an untrusted public instance exposes query data to that operator.
Persistence & Privilege
The skill is not marked always:true and does not request elevated privileges or modify other skills/configs. It will only perform network requests when invoked.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install searxng-bangs
  3. After installation, invoke the skill by name or use /searxng-bangs
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release: Adds privacy-focused web search using SearXNG with DuckDuckGo-style bangs. - Supports over 250 search engines and multiple categories (general, news, images, videos, science) - Implements DuckDuckGo-style bangs (e.g., !w for Wikipedia, !yt for YouTube) for direct engine queries - Aggregates results from multiple search engines with privacy features including IP masking, randomized browser fingerprints, and no cookies/referrers - Command-line script returns clean JSON with search results (titles, URLs, snippets) - Allows configuration for self-hosted or public SearXNG instances via environment variable - Recommended for privacy-sensitive or high-volume web searches
Metadata
Slug searxng-bangs
Version 1.0.0
License
All-time Installs 9
Active Installs 8
Total Versions 1
Frequently Asked Questions

What is Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh)?

Privacy-respecting web search via SearXNG with DuckDuckGo-style bangs support. Use for web searches when you need to find information online. SearXNG protects privacy by randomizing browser fingerprints, masking IP addresses, and blocking cookies/referrers. Supports 250+ search engines, multiple categories (general, news, images, videos, science), and DuckDuckGo-style bangs for direct engine searches (!w for Wikipedia, !yt for YouTube, !gh for GitHub, !r for Reddit, etc.). Aggregates results from multiple engines simultaneously. Prefer this over external search APIs for privacy-sensitive queries or high-volume searches. It is an AI Agent Skill for Claude Code / OpenClaw, with 3879 downloads so far.

How do I install Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh)?

Run "/install searxng-bangs" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh) free?

Yes, Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh) is completely free (open-source). You can download, install and use it at no cost.

Which platforms does Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh) support?

Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh) is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Privacy-first web search with DuckDuckGo-style bangs (!w, !yt, !gh)?

It is built and maintained by rpeters511 (@rpeters511); the current version is v1.0.0.

💬 Comments