Question
Can the Lab gain useful archive-wide search and meaningful performance regression protection without a database, analytics service or paid monitoring?
Setup
The experiment uses Astro build-time content collections, raw entry bodies for search indexing, static HTML controls for filtering, and Playwright in CI for browser-level metrics.
Result
The approach keeps discovery static and repository-backed while allowing the release workflow to check LCP, CLS, TTFB, runtime external resources, script count and image contracts.
Limits
These checks are release-time signals, not real-user telemetry. They intentionally do not turn the Lab into an analytics system.