SearXNG Connect

v1.0.0

Connect to and search the web using SearXNG (privacy-focused meta search engine). No API keys needed - all searches go through your self-hosted SearXNG insta...

0· 287·1 current·1 all-time
byRocco De Angelis@rdeangel

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for rdeangel/searxng-connect.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "SearXNG Connect" (rdeangel/searxng-connect) from ClawHub.
Skill page: https://clawhub.ai/rdeangel/searxng-connect
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Required binaries: uv
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install searxng-connect

ClawHub CLI

Package manager switcher

npx clawhub@latest install searxng-connect
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
Name/description (SearXNG client) match the included Python client and config. Required binary 'uv' is appropriate for running the provided Python script. No unrelated credentials, binaries, or config paths are requested.
Instruction Scope
SKILL.md gives clear, scoped instructions to run the script and explicitly warns about preventing command injection (advises using argv or escaping). It performs web requests to the configured SearXNG instance and optional full‑content fetches of result pages (expected for the purpose). Note: README/CONFIGURE.md reference an execute.sh wrapper that is not present in the file list, so some documented commands may not exist.
Install Mechanism
There is no formal install spec (lowest risk). The README claims the skill 'auto-installs dependencies' and references an execute.sh that would perform setup, but no install script or execute.sh is included — this is a documentation/packaging inconsistency (not evidence of malicious behavior). The script uses the requests library; the skill relies on the system having Python/uv or on the user to install dependencies manually.
Credentials
The skill requests no environment variables or credentials. It reads only its own skill-config.json and writes cache under ~/.openclaw/searxng-cache, which is proportionate to the feature set.
Persistence & Privilege
always:false and no special system-wide modifications. It stores cache/config in its own directories only. Autonomous invocation is allowed by default (normal for skills) but not combined with other alarming privileges.
Assessment
This skill appears to do what it says: it sends search queries to the SearXNG instance you configure and can optionally fetch full page text for richer results. Before enabling it: 1) Ensure you have 'uv' and Python 3.9+ installed (the skill does not include an install script despite README claims). 2) Update skill-config.json to point to a trusted SearXNG instance — all searches go to that URL. 3) Be aware the --full-content option causes the skill to fetch arbitrary result pages (those outbound requests may reveal that your agent requested those URLs and the fetched page content will be visible to the agent); enable it only if you trust that behavior. 4) The SKILL.md correctly warns about command injection; ensure the agent invokes the script using argv-style calls (not raw shell interpolation). 5) If you want the convenience of the referenced execute.sh or an auto-install path, request or add a trustworthy install wrapper rather than running undocumented scripts from unknown sources. Overall this is internally coherent and not disproportionate, but exercise normal caution configuring the target instance and granting autonomous invocation.

Like a lobster shell, security has layers — review code before you run it.

Runtime requirements

🔍 Clawdis
Binsuv
latestvk979z00spjrpfm8ak707wkc6kn82d133
287downloads
0stars
1versions
Updated 1mo ago
v1.0.0
MIT-0

SearXNG Connect

Search the web using your self-hosted SearXNG instance - a privacy-respecting meta search engine that aggregates results from multiple engines without tracking or ads.

AI Usage

When the user asks to search the web, follow this workflow:

Interpret user requests:

  • "Search for [query]" → General web search
  • "Search news about [query]" → Use --categories news
  • "Search images of [query]" → Use --categories images
  • "Search recent [query]" → Use --time-range week or --time-range day
  • "Search [query] last month" → Use --time-range month
  • "Deep search [query]" → Add --full-content for richer page text

Example interpretations:

  • "Search for TypeScript tutorials" → searxng.py "TypeScript tutorials"
  • "Search news about AI" → searxng.py "AI" --categories news --time-range week
  • "Search images of Space Invaders" → searxng.py "Space Invaders" --categories images

Action: Run uv run {baseDir}/scripts/searxng.py "<query>" with appropriate flags.

CRITICAL SECURITY INSTRUCTION: To prevent command injection, do not use raw shell interpolation for the {query} string. Always pass arguments as an argv list: ["uv", "run", "{baseDir}/scripts/searxng.py", query_text, "--categories", "news"]

If your tool only supports a shell string, you MUST properly escape the query input (e.g., using shlex.quote()).


Project

SearXNG - A privacy-focused, self-hosted meta search engine that aggregates results from multiple search engines. It does not track user behavior, does not use cookies, and does not insert ads.

Key Features:

  • Privacy - No tracking, no cookies, no ads - you control the instance
  • Multi-category - Web, news, images, videos, science, IT, files, music, social media
  • Time filtering - Filter results by day, week, month, or year
  • Caching - Local 1-hour cache to avoid redundant requests
  • Rate limiting - Built-in throttling (2 req/sec default)
  • Full content - Optional full page text fetch per result

Requirements

Configuration

Edit skill-config.json to set your SearXNG instance:

{
  "default_instance": "https://your-searxng-instance.com/",
  "cache_enabled": true,
  "cache_expiry": 3600,
  "rate_limit": 2.0
}

Usage

uv run {baseDir}/scripts/searxng.py "TypeScript tutorials"
uv run {baseDir}/scripts/searxng.py "AI news" --categories news --time-range week
uv run {baseDir}/scripts/searxng.py "space photos" --categories images
uv run {baseDir}/scripts/searxng.py "docker tutorial" --no-cache --full-content
uv run {baseDir}/scripts/searxng.py "query" --instance https://my-searxng.com

Options

OptionDescriptionDefault
--categoriesOne or more: general, news, science, files, images, videos, music, social media, itgeneral
--time-rangeyear, month, week, day (aliases: 30d, 7d, 24h, hour)None
--languageBCP-47 language codeen
--pagenoResults page number1
--no-safesearchDisable safe search-
--no-cacheBypass cache for this request-
--full-contentFetch full page text per result (slower, richer)-
--instanceOverride SearXNG instance URLFrom skill-config.json

Troubleshooting

  • Connection refused: Check skill-config.json — verify default_instance URL is correct and SearXNG is running.
  • Empty results: Try a broader query or different category.
  • Slow responses: Remove --full-content or increase rate_limit in config.
  • uv not installed: Install with brew install uv or see https://docs.astral.sh/uv/.

Comments

Loading comments...