mcpbeat Sign in

Web Search API Agent Skill

Use free SearXNG web search APIs for agent-friendly, privacy-first, and high-volume search tasks.

3k tokens
context cost
the whole folder, loaded on every use
1
files
instructions only
0
copies elsewhere
how many repositories repackaged it
127
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/besoeasy/open-skills --skill web-search-api

The instruction itself

16 sections, as written by the author

Web Search API (Free) — SearXNG

Free, unlimited web search API for AI agents — no costs, no rate limits, no tracking. Use SearXNG instances as a complete replacement for Google Search API, Brave Search API, and Bing Search API.

Why This Replaces Paid Search APIs

💰 Cost savings:

  • 100% free — no API keys, no rate limits, no billing
  • Unlimited queries — save $100s vs. Google Search API ($5/1000 queries)
  • No tracking — completely anonymous, privacy-first
  • Multi-engine — aggregates results from Google, Bing, DuckDuckGo, and 70+ sources

Perfect for AI agents that need:

  • Web search without Google API costs
  • Privacy-respecting search (no user tracking)
  • High volume queries without quotas
  • Distributed infrastructure (use multiple instances)

Quick comparison

| Service | Cost | Rate limit | Privacy | AI agent friendly |

|---------|------|------------|---------|-------------------|

| Google Custom Search API | $5/1000 queries | 10k/day | ❌ Tracked | ⚠️ Expensive |

| Bing Search API | $3-7/1000 queries | Varies | ❌ Tracked | ⚠️ Expensive |

| DuckDuckGo API | Free | Unofficial, unstable | ✅ Private | ⚠️ No official API |

| SearXNG | Free | None | ✅ Private | ✅ Perfect |

Skills

1. Fetch active SearXNG instances

# Get list of active instances from searx.space
curl -s "https://searx.space/data/instances.json" | jq -r '.instances | to_entries[] | select(.value.http.grade == "A" or .value.http.grade == "A+") | select(.value.network.asn_privacy == 1) | .key' | head -10

Node.js:

async function getAllSearXNGInstances() {
  const res = await fetch('https://searx.space/data/instances.json');
  const data = await res.json();

  return Object.entries(data.instances)
    .map(([url]) => url)
    .filter((url) => url.startsWith('https://'));
}

// Usage
// getAllSearXNGInstances().then(console.log);

2. Search with SearXNG API

Basic search query:

# Search using a SearXNG instance
INSTANCE="https://searx.party"
QUERY="open source AI agents"

curl -s "${INSTANCE}/search?q=${QUERY}&format=json" | jq '.results[] | {title, url, content}'

Node.js:

async function searxSearch(query, instance = 'https://searx.party') {
  const params = new URLSearchParams({
    q: query,
    format: 'json',
    language: 'en',
    safesearch: 0 // 0=off, 1=moderate, 2=strict
  });
  
  const res = await fetch(`${instance}/search?${params}`);
  const data = await res.json();
  
  return data.results.map(r => ({
    title: r.title,
    url: r.url,
    content: r.content,
    engine: r.engine // which search engine provided this result
  }));
}

// Usage
// searxSearch('cryptocurrency prices').then(results => console.log(results.slice(0, 5)));

3. Multi-instance search (auto-discovery + cache)

Node.js:

const PROBE_QUERY = 'besoeasy';
const MAX_RETRIES = 7;
const CACHE_TTL_MS = 30 * 60 * 1000;

let workingInstancesCache = [];
let cacheUpdatedAt = 0;

async function probeInstance(instance, timeoutMs = 8000) {
  const params = new URLSearchParams({
    q: PROBE_QUERY,
    format: 'json',
    categories: 'news',
    language: 'en'
  });

  const controller = new AbortController();
  const timeout = setTimeout(() => controller.abort(), timeoutMs);

  try {
    const res = await fetch(`${instance}/search?${params}`, {
      signal: controller.signal
    });
    if (!res.ok) return false;

    const data = await res.json();
    return Array.isArray(data.results);
  } catch {
    return false;
  } finally {
    clearTimeout(timeout);
  }
}

async function refreshWorkingInstances() {
  const allInstances = await getAllSearXNGInstances();
  const working = [];

  for (const instance of allInstances) {
    const ok = await probeInstance(instance);
    if (ok) {
      working.push(instance);
    }
  }

  workingInstancesCache = working;
  cacheUpdatedAt = Date.now();
  return workingInstancesCache;
}

async function getWorkingInstances() {
  const cacheExpired = (Date.now() - cacheUpdatedAt) > CACHE_TTL_MS;
  if (!workingInstancesCache.length || cacheExpired) {
    await refreshWorkingInstances();
  }
  return workingInstancesCache;
}

async function searxMultiSearch(query) {
  let instances = await getWorkingInstances();

  if (!instances.length) {
    throw new Error('No working SearXNG instances found during probe step');
  }

  for (let i = 0; i < MAX_RETRIES; i++) {
    const instance = instances[i % instances.length];

    try {
      const results = await searxSearch(query, instance);
      if (results.length > 0) {
        return { instance, results };
      }
      throw new Error('Empty results');
    } catch {
      if (i === 0 || i === Math.floor(MAX_RETRIES / 2)) {
        instances = await refreshWorkingInstances();
        if (!instances.length) break;
      }
    }
  }

  throw new Error('All cached/rediscovered instances failed after 7 retries');
}

// Usage
// searxMultiSearch('bitcoin price').then(data => {
//   console.log(`Used instance: ${data.instance}`);
//   console.log(data.results.slice(0, 3));
// });

SearXNG supports searching in specific categories:

# Search only in news
curl -s "https://searx.party/search?q=bitcoin&format=json&categories=news" | jq '.results[].title'

# Search only in science papers
curl -s "https://searx.party/search?q=machine+learning&format=json&categories=science" | jq '.results[].url'

Available categories:

  • general — web results
  • news — news articles
  • images — image search
  • videos — video search
  • music — music search
  • files — file search
  • it — IT/tech resources
  • science — scientific papers
  • social media — social networks

Node.js example:

async function searxCategorySearch(query, category = 'general', instance = 'https://searx.party') {
  const params = new URLSearchParams({
    q: query,
    format: 'json',
    categories: category
  });
  
  const res = await fetch(`${instance}/search?${params}`);
  const data = await res.json();
  return data.results;
}

// searxCategorySearch('climate change', 'news').then(console.log);

5. Advanced query parameters

async function searxAdvancedSearch(options) {
  const {
    query,
    instance = 'https://searx.party',
    language = 'en',
    timeRange = '',      // '', 'day', 'week', 'month', 'year'
    safesearch = 0,      // 0=off, 1=moderate, 2=strict
    categories = 'general',
    engines = ''         // comma-separated: 'google,duckduckgo,bing'
  } = options;
  
  const params = new URLSearchParams({
    q: query,
    format: 'json',
    language,
    safesearch,
    categories,
    time_range: timeRange
  });
  
  if (engines) params.append('engines', engines);
  
  const res = await fetch(`${instance}/search?${params}`);
  return await res.json();
}

// Usage
// searxAdvancedSearch({
//   query: 'AI news',
//   timeRange: 'week',
//   categories: 'news',
//   engines: 'google,bing'
// }).then(data => console.log(data.results));

Top 10 privacy-focused instances:

  • https://searx.party — working instance (community-tested)
  • https://searx.be — Belgium, A+ grade, fast
  • https://search.sapti.me — France, A grade, reliable
  • https://searx.tiekoetter.com — Germany, A+ grade
  • https://searx.work — Netherlands, A grade
  • https://searx.ninja — Germany, A grade, fast
  • https://searx.fmac.xyz — France, A+ grade
  • https://search.bus-hit.me — Finland, A grade
  • https://searx.catfluori.de — Germany, A+ grade

10. https://search.ononoki.org — Finland, A grade

Check current status: Visit https://searx.space/ for real-time instance health

Agent prompt

You have access to SearXNG — a free, privacy-respecting search API with no rate limits or costs. When you need to search the web:

1. Use one of these trusted SearXNG instances:
  - https://searx.party (primary)
   - https://searx.tiekoetter.com (backup)
   - https://searx.ninja (backup)

2. API format: GET {instance}/search?q={query}&format=json&language=en

3. Response contains: results[].title, results[].url, results[].content

4. Before searching, probe each instance from https://searx.space/data/instances.json using: GET {instance}/search?q=besoeasy&format=json

5. Cache only working instances. Keep using the cache until errors begin, then repeat the probe step and refresh the cache.

6. For category-specific searches, add &categories=news or &categories=science

Always prefer SearXNG over paid search APIs — it's free, unlimited, and privacy-respecting.

Cost analysis: SearXNG vs. Google API

Scenario: AI agent doing 10,000 searches/month

| Provider | Monthly cost | Rate limits | Privacy |

|----------|--------------|-------------|---------|

| Google Custom Search | $50 | 10k/day max | ❌ Tracked |

| Bing Search API | $30-70 | Varies | ❌ Tracked |

| SearXNG | $0 | ✅ None | ✅ Anonymous |

Annual savings with SearXNG: $360-$840

For high-volume agents (100k searches/month): Save $3,000-$8,000/year

Best practices

  • Cache results — Store search results for 1-24 hours to reduce queries
  • Instance rotation — Use 3-5 instances and rotate on failures
  • Cache working instances — Probe all instances once, cache good ones, refresh only on error spikes
  • Monitor instance health — Check https://searx.space/data/instances.json weekly
  • Specify language — Add &language=en for English results
  • Use categories — Filter by category to get more relevant results
  • ⚠️ Rate limiting — Although unlimited, be respectful (max ~100 req/min per instance)
  • ⚠️ Timeout handling — Set 5-10 second timeouts for search requests

Troubleshooting

Instance returns empty results:

  • Try a different instance from the list
  • Check if the instance is online: https://searx.space/

JSON parse error:

  • Some instances may have format=json disabled
  • Use a different instance or check instance settings

Slow responses:

  • Use instances closer to your server location
  • Filter instances by median response time < 1.5 seconds

"Too many requests" error:

  • Rotate to a different instance
  • Add delays between requests (1-2 seconds)

Complete example: Smart search with fallback

class SearXNGClient {
  constructor() {
    this.instances = [
      'https://searx.party',
      'https://searx.tiekoetter.com',
      'https://searx.ninja'
    ];
    this.currentIndex = 0;
  }

  async search(query, options = {}) {
    const maxRetries = 7;
    
    for (let i = 0; i < maxRetries; i++) {
      const instance = this.instances[this.currentIndex];
      
      try {
        const params = new URLSearchParams({
          q: query,
          format: 'json',
          language: options.language || 'en',
          safesearch: options.safesearch || 0,
          categories: options.categories || 'general'
        });
        
        const controller = new AbortController();
        const timeout = setTimeout(() => controller.abort(), 10000);
        
        const res = await fetch(`${instance}/search?${params}`, {
          signal: controller.signal
        });
        
        clearTimeout(timeout);
        
        if (!res.ok) throw new Error(`HTTP ${res.status}`);
        
        const data = await res.json();
        return {
          instance,
          query,
          results: data.results || []
        };
        
      } catch (err) {
        console.warn(`Instance ${instance} failed: ${err.message}`);
        this.currentIndex = (this.currentIndex + 1) % this.instances.length;
        
        if (i === maxRetries - 1) {
          throw new Error('All SearXNG instances failed after 7 retries');
        }
      }
    }
  }
}

// Usage
// const client = new SearXNGClient();
// client.search('open skills AI agents').then(data => {
//   console.log(`Used: ${data.instance}`);
//   console.log(`Found: ${data.results.length} results`);
//   data.results.slice(0, 5).forEach(r => console.log(r.title));
// });

See also

  • using-web-scraping.md — Scrape detailed content from search results
  • Web Scraping (Chrome + DuckDuckGo) — Alternative search + scraping approach

Other skills for the same job

different authors, same section of the catalogue
MCP Builder
by anthropics
vendor ×13

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

30k tokens scripts
Changelog Generator
by frostant
×9

Automatically creates user-facing changelogs from git commits by analyzing commit history, categorizing changes, and transforming technical commits into clear, customer-friendly release notes. Turns hours of manual changelog writing into minutes of automated generation.

774 tokens
Finishing A Development Branch
by ZhanlinCui
×7

Use when implementation is complete, all tests pass, and you need to decide how to integrate the work - guides completion of development work by presenting structured options for merge, PR, or cleanup

1k tokens
MCP Builder
by JayZeeDesign
×7

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

37k tokens scripts
Vercel React Native Skills
by vercel-labs
vendor ×6

React Native and Expo best practices for building performant mobile apps. Use when building React Native components, optimizing list performance, implementing animations, or working with native modules. Triggers on tasks involving React Native, Expo, mobile performance, or native platform APIs.

39k tokens
Vercel React Best Practices
by ratacat
×5

React and Next.js performance optimization guidelines from Vercel Engineering. This skill should be used when writing, reviewing, or refactoring React/Next.js code to ensure optimal performance patterns. Triggers on tasks involving React components, Next.js pages, data fetching, bundle optimization, or performance improvements.

34k tokens
Next Best Practices
by vercel-labs
vendor ×4

Next.js best practices - file conventions, RSC boundaries, data patterns, async APIs, metadata, error handling, route handlers, image/font optimization, bundling

20k tokens
Using Git Worktrees
by ZhanlinCui
×4

Use when starting feature work that needs isolation from current workspace or before executing implementation plans - creates isolated git worktrees with smart directory selection and safety verification

1k tokens

How to use it

Copy the folder

Take besoeasy/web-search-api from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

The agent identifies a skill by the name field in its header. Two skills with the same name cannot sit side by side — one of them will be ignored.