Las instancias de SearXNG con frecuencia son bloqueadas por los motores de búsqueda, lo que hace que los agentes de Hermes pierdan por completo la capacidad de búsqueda. Este tutorial agrega una API alternativa que se activa cuando SearXNG está bloqueado, tiene velocidad limitada o devuelve errores. El agente sigue trabajando independientemente del estado de SearXNG y el respaldo de API cuesta $0,005 por consulta.
Requisitos previos
- Python 3.8+
- solicita biblioteca
- Una clave API de Scavio de scavio.dev
- Agente Hermes con SearXNG configurado
Guia paso a paso
Paso 1: Detectar fallas de SearXNG
Cree un verificador de estado que detecte cuando SearXNG está bloqueado o falla.
Python
import os, requests, time, json
from datetime import datetime
SEARXNG_URL = os.environ.get('SEARXNG_URL', 'http://localhost:8888')
API_KEY = os.environ['SCAVIO_API_KEY']
SH = {'x-api-key': API_KEY, 'Content-Type': 'application/json'}
def check_searxng_health():
"""Check if SearXNG is responsive and returning results."""
try:
resp = requests.get(f'{SEARXNG_URL}/search',
params={'q': 'test', 'format': 'json'}, timeout=5)
if resp.status_code != 200:
return {'healthy': False, 'reason': f'HTTP {resp.status_code}'}
data = resp.json()
if not data.get('results'):
return {'healthy': False, 'reason': 'No results (likely blocked)'}
return {'healthy': True, 'results': len(data['results'])}
except requests.exceptions.ConnectionError:
return {'healthy': False, 'reason': 'Connection refused (SearXNG down)'}
except requests.exceptions.Timeout:
return {'healthy': False, 'reason': 'Timeout (SearXNG overloaded)'}
except Exception as e:
return {'healthy': False, 'reason': str(e)[:50]}
def search_api_fallback(query, num_results=5):
"""Scavio API fallback when SearXNG fails."""
resp = requests.post('https://api.scavio.dev/api/v1/search',
headers=SH, json={'query': query, 'country_code': 'us'}, timeout=10)
resp.raise_for_status()
data = resp.json()
return [{'title': r.get('title', ''), 'url': r.get('link', ''),
'content': r.get('snippet', '')} for r in data.get('organic_results', [])[:num_results]]
health = check_searxng_health()
print(f'SearXNG status: {"healthy" if health["healthy"] else "UNHEALTHY"}')
if not health['healthy']:
print(f' Reason: {health["reason"]}')
print(f' Fallback: Scavio API ready ($0.005/query)')Paso 2: Construya el enrutador de búsqueda inteligente
Enruta las búsquedas a SearXNG cuando esté en buen estado, API cuando no.
Python
class SmartSearchRouter:
def __init__(self):
self.searxng_failures = 0
self.last_check = 0
self.check_interval = 300 # Re-check SearXNG every 5 min
self.api_calls = 0
def search(self, query, num_results=5):
# Try SearXNG first if it has been healthy recently
if self.searxng_failures < 3 or time.time() - self.last_check > self.check_interval:
try:
resp = requests.get(f'{SEARXNG_URL}/search',
params={'q': query, 'format': 'json', 'engines': 'google'},
timeout=5)
if resp.status_code == 200:
data = resp.json()
results = data.get('results', [])
if results:
self.searxng_failures = 0
return {'provider': 'searxng', 'results': [
{'title': r.get('title', ''), 'url': r.get('url', ''),
'content': r.get('content', '')}
for r in results[:num_results]
]}
self.searxng_failures += 1
self.last_check = time.time()
except Exception:
self.searxng_failures += 1
self.last_check = time.time()
# Fallback to API
results = search_api_fallback(query, num_results)
self.api_calls += 1
return {'provider': 'scavio_api', 'results': results}
router = SmartSearchRouter()
# Simulate searches
for q in ['python async tutorial', 'FastAPI deployment', 'database migration tools']:
result = router.search(q)
print(f' [{result["provider"]:12}] {q:30} | {len(result["results"])} results')
print(f'\nAPI fallback calls: {router.api_calls}')
print(f'API cost so far: ${router.api_calls * 0.005:.3f}')Paso 3: Reemplace la búsqueda de Hermes con el enrutador inteligente
Cambie la herramienta de búsqueda Hermes para usar el enrutador inteligente en lugar del SearXNG directo.
Python
def patch_hermes_search():
"""Replace Hermes SearXNG search with smart router."""
router = SmartSearchRouter()
def hermes_search_tool(query, num_results=5):
"""Drop-in replacement for Hermes search tool."""
result = router.search(query, num_results)
# Format in SearXNG response style for compatibility
return {
'results': result['results'],
'number_of_results': len(result['results']),
'_provider': result['provider'],
}
return hermes_search_tool
# Apply the patch
search = patch_hermes_search()
# Test with simulated SearXNG failure
print('Testing with SearXNG down...')
result = search('how to fix SearXNG blocking')
print(f' Provider: {result["_provider"]}')
print(f' Results: {result["number_of_results"]}')
for r in result['results'][:3]:
print(f' {r["title"][:50]}')
print(f'\n SearXNG blocking fix:')
print(f' Problem: SearXNG gets blocked by Google, returns 0 results')
print(f' Solution: Smart router falls back to Scavio API')
print(f' Cost: $0.005/query (only when SearXNG is down)')
print(f' Latency: ~300ms API vs 2-5s SearXNG')Ejemplo en Python
Python
import os, requests
SH = {'x-api-key': os.environ['SCAVIO_API_KEY'], 'Content-Type': 'application/json'}
def search_fallback(query):
"""Use when SearXNG is blocked."""
data = requests.post('https://api.scavio.dev/api/v1/search',
headers=SH, json={'query': query, 'country_code': 'us'}, timeout=10).json()
return [{'title': r['title'], 'url': r['link'], 'content': r.get('snippet', '')}
for r in data.get('organic_results', [])[:5]]
for r in search_fallback('SearXNG alternative'):
print(f'{r["title"][:50]}')
print('SearXNG fallback active. $0.005/query.')Ejemplo en JavaScript
JavaScript
const SH = { 'x-api-key': process.env.SCAVIO_API_KEY, 'Content-Type': 'application/json' };
// Fallback when SearXNG is blocked
const data = await fetch('https://api.scavio.dev/api/v1/search', {
method: 'POST', headers: SH,
body: JSON.stringify({ query: 'SearXNG alternative', country_code: 'us' })
}).then(r => r.json());
(data.organic_results || []).slice(0, 5).forEach(r => console.log(r.title));
console.log('SearXNG fallback ready.');Salida esperada
JSON
SearXNG status: UNHEALTHY
Reason: Connection refused (SearXNG down)
Fallback: Scavio API ready ($0.005/query)
[scavio_api ] python async tutorial | 5 results
[scavio_api ] FastAPI deployment | 5 results
[scavio_api ] database migration tools | 5 results
API fallback calls: 3
API cost so far: $0.015
Testing with SearXNG down...
Provider: scavio_api
Results: 5
SearXNG blocking fix:
Problem: SearXNG gets blocked by Google, returns 0 results
Solution: Smart router falls back to Scavio API
Cost: $0.005/query (only when SearXNG is down)
Latency: ~300ms API vs 2-5s SearXNG