Cornerstone
A semantic logging language. The schema, SDKs, and tooling for events that mean the same thing across boundaries.
Overview
Logging at scale fails for a reason that is not technical. It fails because the events that different services produce do not mean the same thing. Two systems can both emit a `user.login`, but one means a successful authenticated session and the other means a failed credential check. The analyst joining them downstream cannot tell. Schema work is the silent reason that incident response, audit, and analytics become harder than they should be.
Cornerstone is the response. A JSON Schema 2020-12 specification of what an event is, what contexts it can carry, and what shapes those contexts must hold. SDKs in Python and TypeScript, an OpenTelemetry collector, a Go validator, and a VS Code extension are all generated from the same source of truth. When two services emit a Cornerstone event, the analyst joining them downstream has a guarantee that the words mean the same thing.
Highlights
- JSON Schema 2020-12 source of truth. Fourteen context types describe the meaningful shapes events take across teams and products.
- SDKs in Python and TypeScript with a fluent builder API. Strict typing, validation at construction time, consistent shapes across runtimes.
- OpenTelemetry collector processors for schema enforcement, validation, enrichment, and routing. Events flow through the same pipelines as the rest of your telemetry.
- Go CLI for offline validation. VS Code extension with IntelliSense and diagnostics so authors get feedback before events are emitted.