Claude Skill

playwright

Automate browsers and E2E tests with Python Playwright: screenshots, traces.

LLM Mart · 0 points · 2 views 0 listing impressions 0 install-command copies
Download fmind-dot-skills_playwright-3bbe830.zip · 1 KB

Install

skills CLI npx skills add https://github.com/fmind/dot/tree/main/skills/playwright
Claude Code claude plugin marketplace add https://llmmart.ai/marketplace.json && claude plugin install fmind-dot@llmmart
Git git clone https://github.com/fmind/dot.git

The skills CLI installs just this skill, for any of its supported agents. Claude Code installs the whole fmind/dot collection as a plugin from our marketplace. Git is the plain clone.

Skill manifest

Playwright

Browser automation and E2E testing (python-testing owns test structure and fixtures).

Workflow

  1. Check HTTP first: inspect status, headers, and bodies with xh; open a browser only for rendering, interaction, session state, or accessibility, targeting roles/labels and asserting state after each step.
  2. Pin Python packages: uv add --dev playwright pytest-playwright (uv.lock), then uv run playwright install chromium. Report missing OS libraries instead of running privileged install-deps.
  3. Record and capture: uv run playwright codegen --target python-pytest <url>, uv run playwright screenshot <url> <file> (--viewport-size=390,844, --full-page), and uv run playwright pdf <url> <file>. Outside projects that lock Playwright, run uvx --from playwright==<version> playwright (installing chromium first) instead of polluting pyproject.toml.
  4. Run and debug: uv run pytest -q tests/e2e --browser chromium --tracing retain-on-failure --screenshot only-on-failure. Inspect traces in the shell via uv run playwright trace open <trace.zip>, trace --help subcommands, and trace close (show-trace opens the GUI). For live CLI debugging, run one test in the background with --playwright-debug=cli -s, attach via uv run playwright cli attach <session>, then stop the run and remove .playwright-cli/.

Gotchas

  • Respect authorization boundaries: testing never authorizes reusing personal browser profiles/cookies, creating accounts, accepting terms, hitting staging/prod unapproved, or bypassing login/MFA/CAPTCHA.
  • Share ~/.cache/ms-playwright: install only needed browsers and never run playwright uninstall in teardown (other projects share the cache); clean up only temp profiles and passing-run artifacts.
  • Reinstall after version bumps: rerun uv run playwright install chromium whenever Playwright upgrades.

Official Skills

Upstream: microsoft/playwright-cli (CLI exploration only, not Python API); review under the vendor-skill policy. Preview—never install—same-name openai/skills playwright (name collisions). uv run playwright also bundles playwright-cli (uv run playwright cli --help prints its path), playwright-trace, and JS-only playwright-component-testing: audit allowed-tools before exposing (playwright <1.64 and playwright-cli <0.1.22 pre-approve all npx/npm; newer versions scope to npx playwright).

Documentation

Files (dot)
  • SKILL.md 3.6 KB
    ---
    name: playwright
    description: "Automate browsers and E2E tests with Playwright: screenshots and traces."
    license: MIT
    metadata:
      kind: task
      author: Médéric HURIER (Fmind)
      source: github.com/fmind/dot/tree/main/skills/playwright
      created: "2026-09-02"
      updated: "2026-11-10"
    ---
    
    # Playwright
    
    Browser automation and E2E testing ([python-testing](../python-stack/references/python-testing/GUIDE.md) owns test structure and fixtures).
    
    ## Workflow
    
    1. **Check HTTP first**: inspect status, headers, and bodies with [xh](../xh/SKILL.md); open a browser only for rendering, interaction, session state, or accessibility, targeting roles/labels and asserting state after each step.
    1. **Pin Python packages**: `uv add --dev playwright pytest-playwright` (`uv.lock`), then `uv run playwright install chromium`. Report missing OS libraries instead of running privileged `install-deps`.
    1. **Record and capture**: `uv run playwright codegen --target python-pytest <url>`, `uv run playwright screenshot <url> <file>` (`--viewport-size=390,844`, `--full-page`), and `uv run playwright pdf <url> <file>`. Outside projects that lock Playwright, run `uvx --from playwright==<version> playwright` (installing `chromium` first) instead of polluting `pyproject.toml`.
    1. **Run and debug**: `uv run pytest -q tests/e2e --browser chromium --tracing retain-on-failure --screenshot only-on-failure`. Inspect traces in the shell via `uv run playwright trace open <trace.zip>`, `trace --help` subcommands, and `trace close` (`show-trace` opens the GUI). For live CLI debugging, run one test in the background with `--playwright-debug=cli -s`, attach via `uv run playwright cli attach <session>`, then stop the run and remove `.playwright-cli/`.
    
    ## Gotchas
    
    - **Respect authorization boundaries**: testing never authorizes reusing personal browser profiles/cookies, creating accounts, accepting terms, hitting staging/prod unapproved, or bypassing login/MFA/CAPTCHA.
    - **Share `~/.cache/ms-playwright`**: install only needed browsers and never run `playwright uninstall` in teardown (other projects share the cache); clean up only temp profiles and passing-run artifacts.
    - **Reinstall after version bumps**: rerun `uv run playwright install chromium` whenever Playwright upgrades.
    
    ## Official Skills
    
    Upstream: [microsoft/playwright-cli](https://github.com/microsoft/playwright-cli/tree/main/skills/playwright-cli) (CLI exploration only, not Python API); review under the [vendor-skill policy](../agent-project/references/vendor-skills.md). Preview—never install—same-name `openai/skills` `playwright` ([name collisions](../agent-project/references/vendor-skills.md#name-collisions)). `uv run playwright` also bundles `playwright-cli` (`uv run playwright cli --help` prints its path), `playwright-trace`, and JS-only `playwright-component-testing`: audit `allowed-tools` before exposing (`playwright <1.64` and `playwright-cli <0.1.22` pre-approve all `npx`/`npm`; newer versions scope to `npx playwright`).
    
    ## Documentation
    
    - [Playwright for Python](https://playwright.dev/python/docs/intro) · [pytest plugin](https://playwright.dev/python/docs/test-runners) · [Trace Viewer](https://playwright.dev/python/docs/trace-viewer)
    - Releases: [Playwright release notes](https://playwright.dev/python/docs/release-notes) · [playwright-python](https://github.com/microsoft/playwright-python/releases)
    - Companion skills: [python-stack](../python-stack/references/foundation/GUIDE.md), [chrome-devtools](../chrome-devtools/SKILL.md) (accessibility, Lighthouse, and performance audits), [performance](../python-stack/references/performance.md) (load, not browser, testing).
    

Comments (0)

Sign in to join the conversation.

No comments yet.

Reviews (0)

No reviews yet.