Maple
Open-source traces, logs and metrics on OpenTelemetry, queryable by an AI agent.
Maple ingests traces, logs and metrics over standard OpenTelemetry into a ClickHouse backend, running sub-second queries across billions of rows — currently 41.5B+ traces a month. Every surface shares one trace id, so an alert links straight to the failing span, its correlated logs, and a session replay, and its MCP server lets Claude, Cursor or any MCP client search traces, read the source and propose the fix itself. It's open source (FSL-1.1, converting to Apache 2.0 two years after each release) and self-hostable, so telemetry never has to leave your own infrastructure.
What holds up
- +One trace id links the alert, the failing span, logs and session replay — no stitching.
- +Self-hostable against your own ClickHouse — telemetry never has to leave your infra.
- +MCP server lets Claude or Cursor search traces, read source and propose a fix directly.
Mind the limits
- −Source is FSL-1.1, not fully open — becomes Apache 2.0 only two years after each release.
- −Self-hosting means running and scaling the ClickHouse backend yourself.
Featured here? Take the badge
Put it on your site — it links back to this review. Free for every listed product, always.
<a href="https://stillworks.dev/products/p/maple/"><img src="https://stillworks.dev/badge/maple.svg" alt="Picked by StillWorks" width="250" height="54"></a>