Changelog is doubling down on AI agent integration and observability parity. The platform added dedicated skills that let AI agents write evaluations and query metrics directly, graduating metrics to general availability with hyper-cardinality support unified alongside logs and traces. A programmatic Dashboards API shipped alongside online evaluations that score AI outputs against live production traffic, positioning the platform as an end-to-end system for building and validating AI capabilities.
Changelog unified observability with AI agent capabilities across the month. Metrics graduated to GA with hyper-cardinality support alongside logs and traces, queryable by agents through MCP and a dedicated skill. The team also shipped a Dashboards API for programmatic control and a Write Evaluations skill that lets agents author test suites.