Skip to main content
Agent skillFirst-run guide

Webapp Testing

Use Python Playwright to test local web applications, inspect visible UI behavior, capture screenshots, and collect browser logs.

Get started

Get the complete skill folder, including its scripts and reference files.

·

When to use this skill

Use Web App Testing to exercise a local web app through its browser controls: forms, navigation, dynamically rendered results, and visible empty states. The Anthropic skill uses Python Playwright and includes a server lifecycle helper.

When not to use it

Use narrower tests for pure data helpers or calculations. A browser smoke test alone does not prove that a production deployment, every browser, or every accessibility requirement works.

How to install

Where will you use it?

Install the publisher’s plugin

This installs a bundle that can include other skills. Choose this method or the manual alternative below; you only need one installation.

  1. Run these commands in Claude Code:

    Claude Code CLI
    /plugin marketplace add anthropics/skills
    /plugin install example-skills@anthropic-agent-skills
  2. Ask Claude Code to use Webapp Testing on a small task and confirm that it loads. Check the setup requirements in the publisher’s guide if it cannot run.
Alternative: install the folder manually

Use these steps if you choose the manual folder method. If you installed the plugin above, skip this alternative.

  1. Claude Code personal install: place the skill and its bundled files at ~/.claude/skills/webapp-testing/SKILL.md. For a shared project, commit the folder at .claude/skills/webapp-testing/ instead.
  2. Current Claude Code detects SKILL.md changes in watched directories during the session. If you just created a top-level skills directory, run /reload-skills. Check that the skill is listed before testing its trigger.
  3. For Claude chat or Cowork in the Desktop app, manage uploaded skills in Customize > Skills; copying files into a local Claude Code folder is not a Desktop install. Other agents may use different install locations.

Full install guide for Claude, Cursor, and Codex →

What supports this recommendation?

81/100 provisional content score. This measures documentation, practical detail and source signals. It is not a measured task success rate or a security certification.

Documentation
85/100
Practical detail
76/100
Source evidence
70/100
Source signals
96/100
Model compatibility evidence
  • claudeuntested

    No model-specific signal or recorded compatibility test was found.

  • chatgptuntested

    No model-specific signal or recorded compatibility test was found.

  • geminiuntested

    No model-specific signal or recorded compatibility test was found.

  • copilotuntested

    No model-specific signal or recorded compatibility test was found.

  • llamauntested

    No model-specific signal or recorded compatibility test was found.

  • perplexityuntested

    No model-specific signal or recorded compatibility test was found.

  • mistraluntested

    No model-specific signal or recorded compatibility test was found.

  • grokuntested

    No model-specific signal or recorded compatibility test was found.

Inferred compatibility is based on content signals. Check the app, its available tools and setup requirements before running a skill.

Source reviewedExample execution recorded

Try one small task

Best first exercise: verify one visible browser journey and save its evidence.

See actual outputs and execution limits →

What you need

  • A locally running web application, Python Playwright, and a browser installed for Playwright.
  • The complete Anthropic webapp-testing skill folder; its server helper is useful when the app is not running yet.

Check a local search box with a known match and a no-results case.

First task prompt
Use webapp-testing on my local app. First ask me for its local URL and one known searchable item if those are missing. In Playwright, submit that item and verify a visible matching result, then submit a unique nonsense query and verify the empty state. Save screenshots and browser console errors. Use the actual page controls, report what passed or failed, and do not submit forms that create or delete data.

Expected result

A specification to check against, not output from an AI run.

A repeatable browser test, two screenshots, and observed pass/fail results for the two search states.

  • The test interacts with the visible search UI and checks results after navigation or rendering.
  • The matching query shows the supplied known item; the nonsense query shows an explicit empty state.
  • Screenshots and console observations come from the executed browser session.

Where this exercise stops

  • This Anthropic skill is distinct from the Superpowers testing workflows.
  • A local smoke test does not establish production accessibility, browser coverage, or performance.

Recorded example run

These files came from actual execution in a Codex session. They were produced and checked within the same session; this is scoped evidence, not independent review or a measured success rate.

Search tested in a real browser

Executed · Automated Codex exercise

Codex session; Python Playwright; local Next.js preview; headless Chromium 147.0.7727.0

The agent followed the pinned webapp-testing instructions with Python Playwright against a local preview, used visible search controls, and saved browser screenshots.

Download the actual outputs

Recorded checks and observations

  • Passed: Known PDF query entered through visible search controls
  • Passed: Nonsense query shows an explicit empty state and no skill cards
  • Passed: Screenshots saved from the executed browser session
  • Passed: No uncaught browser JavaScript errors
  • Passed: No browser console errors
Scope, limitations and pinned sources
  • The agent read the pinned upstream instructions and used their prescribed tools. Native Claude skill loading and general cross-agent compatibility were not tested.
  • One local Chromium run, not production or cross-browser coverage.
  • Third-party analytics and product telemetry were intercepted; no synthetic demand was recorded.
  • This is an automated Codex exercise, not an independent human evaluation or no-skill comparison.
Read the execution record (JSON) →
Source and evidence

The upstream instructions informed this guide. The exercise and acceptance checks are an OpxSkills test recipe. Where a run is recorded, its actual outputs, environment and limits appear separately above. A recorded example does not establish general model compatibility or a success rate.

Source reviewed · Upstream revision 8a1541c4a3ff

What this skill does

This Anthropic skill provides a workflow for testing local web applications with Python Playwright. It includes a helper for starting the required servers, plus examples for capturing screenshots and inspecting browser behavior.

Intended uses

  • Verify a signup form error state
  • Catch a page that only renders after JavaScript
  • Regression-check a checkout flow

Implementation notes

  • Assert visible text, not class names.
  • Prefer the same path a user takes over hitting internals.

Try it, then tell us how it went

A copied prompt does not tell us whether the task worked. Report your result after checking the output. We collect your tool and outcome, never your files or prompt.