Triage Ci Flake

This handles the tedious workflow of debugging flaky CI tests that mysteriously fail on main but pass locally. When you point it at a failing test or PR, it systematically reproduces the issue by running the specific test against both dev and production builds, not just reading logs and guessing. The PR aggregation feature is clever: it pulls all failing checks from a PR number, extracts the actual test names and errors, then patterns across them to spot if it's a selector change breaking multiple suites or isolated flakiness. The source is adamant about mandatory reproduction steps before any fixes, which is the right instinct. Worth using when you're staring at a red CI check wondering if it's timing, bundling, or test pollution causing the failure.