Observability
What you can see about your MCPify service, and how to get at it
MCPify runs your service on a hosted gateway, so the telemetry lives with us rather than in your infrastructure. This page says plainly what exists today and what does not — a question worth answering before you design around it.
What is collected today
- Request logs for your service, retained in Google Cloud Logging
- Gateway-side latency and error-rate metrics per service
- Token accounting per tool call, which is how response shaping and truncation are tuned
- Cache behaviour per tool — what is cached, for how long, and what that changes
Ask and we will pull what you need for your service — traffic volumes, error patterns, latency distribution, or how a specific tool is behaving. For an ongoing arrangement, say what you want to see and how often, and we will work out how to deliver it.
What does not exist yet
- —A self-serve analytics dashboard
- —A customer-facing analytics or reporting API
- —Scheduled report delivery
- —Direct connectors to Datadog, Tableau or similar
If one of these would decide an integration for you, tell us — knowing which of them actually matters to customers is how it gets built.