About Spanly

Spanly is an observability platform purpose-built for the Model Context Protocol. We help engineering teams shipping MCP in production monitor, debug, and operate their servers with real-time tracing, error tracking, and performance metrics.

Our customers are typically B2B SaaS companies that have added an MCP server to their existing product. They already run a general-purpose APM. What they need is a layer that understands the MCP protocol itself (tool calls, prompts, resource reads) and the people on call to read them at 3am.

Our mission

Datadog, Sentry, and New Relic do an excellent job on HTTP and infrastructure. They don't model MCP. Spanly adds the MCP-protocol layer on top (request semantics, tool calls, prompts, and resource reads), so you can keep the observability stack you already run and add a dedicated lens for the new surface.

Built for engineering teams

Spanly is built for the people who carry the pager. Single sign-on, an audit log of every action, alert routing through Slack and PagerDuty, and a public status board for the office TV. The SDK is open source and adds under a millisecond per request; the CLI exports to any OpenTelemetry-compatible APM.