| name | webapp-testing |
| description | Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs. |
| emoji | 🧪 |
Prefer small, native Playwright (Python) scripts over ad-hoc clicking.
Recon-then-action pattern:
- First inspect the rendered DOM — take a screenshot and dump the content.
- Identify selectors from what you actually see (not what you assume).
- Then act on those selectors.
Decision: static HTML -> just read it; dynamic content -> drive a real browser.
Pitfalls: don't inspect the DOM before the app finishes loading — wait for
networkidle on dynamic apps. Use the synchronous Playwright API, close
browsers properly, use descriptive selectors, and add appropriate waits.
Capture console logs to debug runtime errors.
Source: github.com/anthropics/skills (Apache-2.0). Condensed + attributed for the corlinman marketplace; see ../../ATTRIBUTION.md.