Playwright Issue Triage
Reproduce and verify Playwright GitHub issues, then report a defensible status.
The skill requires reading GitHub issues, cloning external repositories, and running reproduction code. It does warn the user to inspect suspicious code first and avoid fixes or misrouting, but it lacks explicit confirmation, sandbox isolation, credential/sensitive-data handling, network boundaries, and rollback guidance, so substantial points are deducted.
The classification, reproduction, version fallback, cross-browser checks, and reporting flow are coherent, and it requires documenting failed reproduction attempts. However, it depends on browsers, Node, network access, and external repositories; its advice to guess missing inputs lacks precise prerequisites and failure diagnostics. Static review therefore caps this below 10.
Triggers, issue categories, and the non-fit boundary around fixing are reasonably clear, and the skill correctly covers Playwright family projects. It does not define Chinese-language output or mainland-China reachability, and it heavily depends on GitHub, npm, browser downloads, and a suitable local checkout, so boundary and environment coverage are incomplete.
The document has clear sections for classification, reproduction, test reduction, reporting, and safety notes, with repository examples. It lacks skill-specific installation and dependency notes, a stable output template, known limitations, compatibility/version policy, changelog, and explicit maintenance ownership, so it cannot approach full marks.
The intended result is a clear, verified issue status, and the workflow can guide a human triager directly. However, completion requires substantial manual execution, external inputs, and environment setup; the skill provides no fixed machine-consumable output format or statically verifiable result. The score remains conservative under the static cap.
Repository test examples, bot-voice guidance, and prescribed version/source checks provide some auditability. There are no skill-specific committed tests, CI coverage, or third-party execution artifacts, and no steps were executed in this review, so only limited credit is justified.
- Obtain user confirmation before running code from a linked issue repository, isolate the workspace, restrict network and credential access, and prevent npm scripts or tests from reaching production data.
- Cloning repositories, downloading browsers, and testing multiple versions may fail because of GitHub, npm, or browser-CDN reachability; provide explicit offline or proxy-failure fallbacks.
- Add a stable report template, timeout and exception handling, sensitive-log redaction, and recoverable cleanup rules for temporary directories.
What does this skill do, and when should you use it?
This skill analyzes Playwright GitHub issues and classifies them as bugs, feature requests, upstream or environment issues, or usage questions. For bugs, it reads the full thread, extracts versions, browsers, operating systems, and reproduction details, then tests across browsers and relevant variations. It can condense a complex reproduction into a self-contained Playwright test. The result is a maintainer-style report with status, evidence, test coverage, and a minimized reproduction—not a fix.
Reads GitHub issues and comments; collects versions, browsers, operating systems, reproduction repositories or snippets, and expected-versus-actual behavior; reviews linked code before running it; tests tip-of-tree first and then the reported version; exercises Chromium, Firefox, and WebKit in headed/headless and other relevant variations; bisects versions when appropriate; writes a single self-contained test under tests/page/ and runs npm run ctest; produces a status such as reproduced, fixed-on-latest, cannot-reproduce, or not-a-bug, with supporting evidence.
- A Playwright maintainer receives a bug report with a minimal reproduction and needs to verify whether it still occurs.
- An engineer reports behavior that appears limited to one browser and needs a Chromium, Firefox, and WebKit comparison.
- A maintainer needs to reduce an application-specific reproduction to a repository-style self-contained test.
- A team needs to determine whether an issue belongs to Playwright, a browser engine, an external website, or usage/documentation.
What are this skill's strengths and limitations?
- Provides explicit classification across bugs, feature requests, upstream/environment issues, and usage questions.
- Requires thorough testing across Chromium, Firefox, WebKit, and relevant execution modes.
- Promotes self-contained reproductions using Playwright's existing test fixtures.
- Reports the complete test matrix, browser-specific divergence, and evidence.
- Depends on access to GitHub issues, linked repositories, and the network.
- Requires shell and filesystem access plus a Node.js, npm, Git, and Playwright test environment.
- The source does not document a standalone installation flow or an included automated test suite.
- The scope ends at triage, reproduction, and status reporting rather than fixing the issue.
How do you install this skill?
The source does not document a standalone installation command for this skill. The skill file is located at .claude/skills/playwright-triage/SKILL.md. For a Playwright test environment, the README supports npm i -D @playwright/test followed by npx playwright install.
How do you use this skill?
Ask a compatible client with a request such as: "Triage and reproduce GitHub issue #123 in microsoft/playwright." Provide the issue URL or number, the full discussion, and any available reproduction repository or snippet. The skill triages and reports the issue; it does not implement a fix.