AMI.
Estado del sistema System status
Todos los sistemas operativos All systems operational
Última verificación: hace un instante · auto-refresh cada 60 s Last check: just now · auto-refresh every 60 s

Componentes Components Estado en vivo de cada superficie Live state of each surface

API REST REST API
https://api.protocolami.com
Operativo Operational
Uptime 90d
99.99%
p50
12 ms
p99
47 ms
Sin incidentes en 90 días. No incidents in 90 days.
MCP HTTP MCP HTTP
https://mcp.protocolami.com/mcp/
Operativo Operational
Uptime 90d
99.97%
p50
18 ms
p99
62 ms
Sin incidentes en 90 días. No incidents in 90 days.
Persistencia (SQLite) Persistence (SQLite)
internal · file:ami.db
Operativo Operational
Uptime 90d
100.00%
p50
0.4 ms
p99
2.1 ms
Sin incidentes en 90 días. No incidents in 90 days.
Adapter telco (mock) Telco adapter (mock)
internal · ami_telco.mock
Operativo Operational
Uptime 90d
100.00%
p50
1 ms
p99
3 ms
En desarrollo activo · adapter de referencia. In active development · reference adapter.
Partner telco (live) Telco partner (live)
pending · live SIM provisioning
Degradado Degraded
Uptime 90d
p50
p99
Pendiente de activación con el partner telco. Awaiting partner activation.
MCP server (stdio) MCP server (stdio)
local · pip install ami-mcp
Operativo Operational
Uptime 90d
100.00%
p50
1 ms
p99
4 ms
Sin incidentes en 90 días. No incidents in 90 days.
Documentación Documentation site
https://protocolami.com/docs
Operativo Operational
Uptime 90d
99.98%
p50
9 ms
p99
31 ms
Sin incidentes en 90 días. No incidents in 90 days.

Uptime Uptime Uptime histórico Historical uptime

Uptime histórico (90 días) Historical uptime (90 days)
verde = OK · amarillo = degradado green = OK · amber = degraded
90d60d30dtoday

Incidentes Incidents Historial de incidentes Incidents history

2026-04-14 · resuelto 2026-04-14 · resolved
Latencia elevada en MCP HTTP (28 min) Elevated latency on MCP HTTP (28 min)

Un cliente abrió ~3.000 sesiones MCP en paralelo desde un único IP, agotando el pool de workers. Mitigado con un rate-limit por origen; rooted-cause cerrada con un cap global de sesiones concurrentes por cuenta. A client opened ~3,000 parallel MCP sessions from a single IP, exhausting the worker pool. Mitigated with a per-origin rate-limit; root cause closed with a global cap on concurrent sessions per account.

Auto-refresh cada 60 s · Status público mantenido por Parallax IEI Auto-refresh every 60 s · Public status maintained by Parallax IEI JSON crudo /v1/health Raw JSON /v1/health