Searxng Search

SearXNG Search

Use scripts/search.py to perform web searches and return structured results.

Configuration

Configuration is resolved in the following priority order:

1. Environment variables (highest priority)
2. .env file in the current working directory (auto-loaded if present)
3. Built-in defaults

| Env var | Purpose | Default | |-----------------|----------------------------------|--------------------------------| | SEARXNG_URL | Base URL of SearXNG instance | https://search.981234.xyz | | SEARXNG_TOKEN | Bearer token for auth (optional) | (empty = no auth header sent)|

Example .env:

SEARXNG_URL=https://search.example.com
SEARXNG_TOKEN=your-secret-token

Usage

# Basic search (uses defaults)
python3 scripts/search.py "query"

# Paginate or limit
python3 scripts/search.py "query" --page 2 --limit 5

# Custom instance with auth
SEARXNG_URL=https://my.instance.com SEARXNG_TOKEN=my-token python3 scripts/search.py "query"

Output format

NDJSON — one JSON object per line:

{"url": "https://example.com", "title": "Page Title", "snippet": "Relevant excerpt..."}

Workflow

1. Run the script, capturing stdout.
2. Parse NDJSON lines into a list of results.
3. Present to the user as a numbered list of links with snippets.
4. If the user wants more results, re-run with --page N or a higher --limit.

Notes

• Default limit is 10; one server page typically returns ~40 results.
• If SEARXNG_TOKEN is unset or empty, the Authorization header is omitted (public instances).
• Script exits with code 1 and prints to stderr on HTTP or network failure.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: ichuan
• Source: ichuan/skills
• License: MIT

Install and usage instructions live in the source repository linked above.