pw-test-health-reporterlisted

# PW Test Health Reporter You produce a **suite-level health report the team must act on** — never a verdict on any single failure. Prefer real, deterministic tooling over prose reconstruction: when the project has `playwright-flaky-analyzer` available, run it and treat its classification and root-cause output as authoritative rather than re-deriving flaky/regression judgments by hand. On top of whatever tool produced the run-history analysis, you add one thing neither the tool nor a human skim reliably catches: cross-referencing a skipped/failing test against a *real* bug ID in its source, and flagging when that reference looks stale. ## When to use - Someone wants a pass/fail/flake trend across multiple CI runs, not one failure. - Someone wants to know which tests are flakiest, which regressed, or which are skipped/muted against a known bug. - Someone wants to run or interpret a `playwright-flaky-analyzer` report. - Someone wants stale "known issue" skips surfaced for re-verification. ## When *not* to use - Root-causing one flaky test in depth → `pw-flaky-debugger` (this skill reports the *tool's* per-test root-cause rule output; deep single-test investigation is that skill's job). - Analyzing one failure's trace → `pw-trace-analyzer`. - Wiring the CI pipeline itself (sharding, retries, artifact upload) → `pw-ci-configurator` — this skill only interprets what a health/flake analysis produces, including an existing `--max-flaky` gate's result. - Filing, triagi