Save

Five testing workflows in one skill: scan for high-value test candidates, write focused tests, generate realistic workloads with adversarial edge cases, review whether existing tests are worth keeping, and diagnose test suite health. Built around a reference library for risk-based testing, test level selection, oracles and assertions, and workload modeling. The stance is opinionated: protect meaningful behavior over coverage numbers, ask what could go wrong before writing tests, and keep a test only if you can name a plausible bug it would catch. Commands route through $wio scan, test, workload, review, or doctor. Worth trying if you're drowning in low-signal tests or keep asking "what should I even test here?"

Install to Claude Code

npx -y skills add workersio/spec --skill save --agent claude-code