Resumen
Meeting notes generar accion elementos ese reference herramientas, precios, y competidores. Teams act on stale informacion de el meeting sin verifying. Este flujo de trabajo toma meeting accion elementos y automaticamente grounds them con live search datos via MCP, checking si referenced herramientas aun exist, precios tiene changed, o competidores tienen nuevo offerings.
Desencadenador
Activado despues de meeting transcript es procesado, o diario batch for accumulated accion elementos.
Programación
Post-meeting o diario batch
Pasos del flujo de trabajo
Extraer Accion Items de Meeting Notes
Analizar meeting transcript o notes for accion elementos ese reference external herramientas, precios, competidores, o factual claims.
Generar Search Consultas per Accion Item
For cada accion elemento, generar 1-3 consultas de busqueda to verify el referenced informacion. Precios claims obtener precios consultas. Herramienta references obtener disponibilidad consultas.
Ejecutar Grounding Searches via MCP
Ejecutar todos generado consultas de busqueda via el Scavio MCP servidor. Recopilar structured resultados con marcas de tiempo for freshness.
Annotate Accion Items con Findings
Agregar cada accion elemento con relevante search findings. Marcar elementos donde el referenced informacion tiene changed desde el meeting.
Implementacion en Python
import requests, os
API_KEY = os.environ["SCAVIO_API_KEY"]
def ground_action_item(action_item: str) -> dict:
"""Ground a meeting action item with live search data."""
resp = requests.post(
"https://api.scavio.dev/api/v1/search",
headers={"x-api-key": API_KEY, "Content-Type": "application/json"},
json={"query": action_item, "country_code": "us"},
timeout=10,
)
data = resp.json()
return {
"action_item": action_item,
"grounding": [
{"title": r.get("title", ""), "snippet": r.get("snippet", ""), "url": r.get("link", "")}
for r in data.get("organic_results", [])[:3]
],
"ai_overview": data.get("ai_overview", {}).get("text", ""),
}
# Ground meeting action items
items = [
"Evaluate Vercel pricing for deployment migration",
"Check if Supabase supports edge functions in free tier",
]
for item in items:
grounded = ground_action_item(item)
print(f"Action: {grounded['action_item']}")
print(f" Grounding: {len(grounded['grounding'])} sources")
if grounded["ai_overview"]:
print(f" AI Overview: {grounded['ai_overview'][:100]}...")Implementacion en JavaScript
const H = {'x-api-key': process.env.SCAVIO_API_KEY, 'Content-Type': 'application/json'};
async function groundActionItem(item) {
const r = await fetch('https://api.scavio.dev/api/v1/search', {method:'POST', headers:H, body:JSON.stringify({query:item, country_code:'us'})});
const d = await r.json();
return {action_item:item, grounding:(d.organic_results||[]).slice(0,3).map(r=>({title:r.title, snippet:r.snippet, url:r.link})), ai_overview:d.ai_overview?.text||''};
}
const items = ['Evaluate Vercel pricing for deployment migration', 'Check if Supabase supports edge functions'];
for (const item of items) {
const g = await groundActionItem(item);
console.log('Action: '+g.action_item+'\n Sources: '+g.grounding.length);
}Plataformas utilizadas
Búsqueda web con grafo de conocimiento, PAA y resúmenes de IA