Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Browser automation via Playwriter (remorses) using persistent Chrome sessions and the full Playwright Page API.
.claude/skills/mikeyobrien-playwriter/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -21% | 0% |
Use Playwriter to run Playwright Page scripts against your local Chrome session. This preserves logins, cookies, and extensions, which is ideal for web dashboard testing and authenticated flows.
bash npm i -g playwriter
bash playwriter session new
bash playwriter session list
bash playwriter -s <session_id> -e "await page.goto('https://example.com')"
Within -e, these are available in scope:
page (Playwright Page)context (BrowserContext)state (persistent object across calls in the same session)require (for loading helper modules)Example state persistence:
bashplaywriter -s <session_id> -e "state.lastUrl = page.url()" playwriter -s <session_id> -e "console.log(state.lastUrl)"
bashplaywriter -s <session_id> -e "await page.goto('http://localhost:3000'); await page.getByRole('button', { name: 'Run' }).click();"
bashplaywriter -s <session_id> -e "await page.getByLabel('Email').fill('qa@example.com'); await page.getByLabel('Password').fill('secret'); await page.getByRole('button', { name: 'Sign in' }).click();"
bashplaywriter -s <session_id> -e "const { screenshotWithAccessibilityLabels } = require('playwriter'); await screenshotWithAccessibilityLabels(page, { path: '/tmp/a11y.png' });"
bashplaywriter -s <session_id> -e "await page.route('**/api/**', async route => { const res = await route.fetch(); const body = await res.json(); await route.fulfill({ json: { ...body, injected: true } }); });"
bashplaywriter -s <session_id> -e "const text = await page.locator('main').innerText(); console.log(text);"
getByRole and getByLabel for stable selectors.state.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 10,400 | 9,242 | -11% | 1 | 1 | 0% | 2,185 | 2,862 | +31% | 0 | 0 | — |
case-22 | pass→pass | 7,155 | 6,840 | -4% | 1 | 1 | 0% | 1,605 | 2,098 | +31% | 0 | 0 | — |
case-01 | fail→fail | 7,939 | 1,908 | -76% | 1 | 1 | 0% | 1,672 | 937 | -44% | 0 | 0 | — |
case-02 | fail→pass | 12,422 | 13,157 | +6% | 1 | 1 | 0% | 1,211 | 1,319 | +9% | 0 | 0 | — |
case-03 | fail→pass | 9,315 | 3,979 | -57% | 1 | 1 | 0% | 1,903 | 1,596 | -16% | 0 | 0 | — |
case-04 | pass→pass | 8,203 | 2,449 | -70% | 1 | 1 | 0% | 1,677 | 910 | -46% | 0 | 0 | — |
case-05 | fail→pass | 8,275 | 3,692 | -55% | 1 | 1 | 0% | 1,534 | 970 | -37% | 0 | 0 | — |
case-06 | fail→pass | 8,574 | 2,998 | -65% | 1 | 1 | 0% | 1,722 | 1,138 | -34% | 0 | 0 | — |
case-07 | fail→pass | 7,210 | 2,022 | -72% | 1 | 1 | 0% | 1,363 | 1,078 | -21% | 0 | 0 | — |
case-08 | fail→pass | 11,731 | 2,222 | -81% | 1 | 1 | 0% | 2,242 | 1,084 | -52% | 0 | 0 | — |
case-09 | fail→pass | 11,539 | 2,956 | -74% | 1 | 1 | 0% | 2,370 | 1,350 | -43% | 0 | 0 | — |
case-10 | pass→pass | 14,188 | 9,462 | -33% | 1 | 1 | 0% | 2,571 | 2,484 | -3% | 0 | 0 | — |
case-11 | fail→pass | 8,635 | 2,381 | -72% | 1 | 1 | 0% | 1,448 | 1,087 | -25% | 0 | 0 | — |
case-12 | fail→pass | 8,305 | 888 | -89% | 1 | 1 | 0% | 1,463 | 809 | -45% | 0 | 0 | — |
case-13 | fail→pass | 13,127 | 2,485 | -81% | 1 | 1 | 0% | 2,413 | 1,089 | -55% | 0 | 0 | — |
case-14 | fail→pass | 7,803 | 2,752 | -65% | 1 | 1 | 0% | 1,469 | 1,291 | -12% | 0 | 0 | — |
case-15 | fail→pass | 6,328 | 2,003 | -68% | 1 | 1 | 0% | 1,212 | 1,045 | -14% | 0 | 0 | — |
case-16 | pass→pass | 11,434 | 5,268 | -54% | 1 | 1 | 0% | 2,215 | 1,670 | -25% | 0 | 0 | — |
case-17 | fail→pass | 10,178 | 3,826 | -62% | 1 | 1 | 0% | 1,983 | 1,378 | -31% | 0 | 0 | — |
case-18 | fail→pass | 7,840 | 1,465 | -81% | 1 | 1 | 0% | 1,601 | 938 | -41% | 0 | 0 | — |
case-19 | fail→pass | 14,461 | 1,732 | -88% | 1 | 1 | 0% | 2,642 | 1,067 | -60% | 0 | 0 | — |
case-20 | pass→pass | 12,426 | 7,089 | -43% | 1 | 1 | 0% | 2,895 | 2,349 | -19% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +68 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.