agentleFS
MicrosoftKnown publishermicrosoft/playwright/.claude/skills/playwright-audit-deps/SKILL.md

playwright-audit-deps

Skill100k starsChanged 16 days agoApache-2.0View on GitHub

Roll the runtime dependencies that npm audit flags with a moderate or higher advisory to safe versions that are at least 8 days old, and commit them with a table of every package the lock file change touched (old and new version, publish date, weekly downloads). Use when asked to fix audit findings, bump vulnerable or outdated runtime dependencies, or update package-lock.json for an advisory.

  • Installs packages

Trending

More talk lately.about the whole repository

Dec 2025: 3 mentionsJan 2026: 3 mentionsFeb 2026: 2 mentionsMar 2026: 2 mentionsApr 2026: 1 mentionMay 2026: 1 mentionJun 2026: 0 mentionsJul 2026: 0 mentionsAug 2026: 1 mentionSep 2026: 3 mentionsOct 2026: 4 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the microsoft/playwright repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about microsoft/playwright

The biggest 5 of 119 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 119

Talked about on Bluesky, GitHub, Stack Overflow, Hacker News, the web and YouTube (119 times).

On Bluesky · 39

34 more on Bluesky

On GitHub · 31

26 more on GitHub

On Stack Overflow · 25

20 more on Stack Overflow

On the web · 9

4 more on the web

On YouTube · 6

1 more on YouTube

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Start from a clean checkout of main with npm run build done, so the bundle sidecars exist.

    line 26
  2. Very specific:

    Verify: npm run build, then npm run itest tests/installation/bundle-licenses.spec.ts tests/installation/bundle-size.spec.ts and npm run …

    line 38
  3. Very specific:

    Commit package.json and package-lock.json as one commit:

    line 43
  4. Distinctive:

    Run the apply step once from the clean base.

    line 64

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Roll runtime dependencies flagged by npm audit

/playwright-audit-deps [issue url]

A runtime dependency is a package whose code ends up in a published Playwright package: it is inlined into a bundle (listed in a packages/*/lib/**/*.js.txt sidecar) or named in the dependencies of a packages/*/package.json. Advisories on dev-only packages, such as eslint plugins or chromium-bidi, are reported as skipped and left alone.

A version is eligible when it is outside every advisory range, was published at least 8 days ago, and is accepted by every runtime dependent in package-lock.json. The newest eligible version wins. Packages that npm install pulls in beyond that age are pinned back the same way.

The lint npm run check-dependency-age (part of npm run lint and npm run flint, so CI runs it on every PR) fails when any version in package-lock.json is less than 7 days old. The 8-day default here keeps a one-day margin above that limit.

Steps

  1. Start from a clean checkout of main with npm run build done, so the bundle sidecars exist. Create the branch: fix-<issue number> when an issue is given, else bump-runtime-deps.
  2. Plan: node .claude/skills/playwright-audit-deps/audit-deps.mjs. It prints, per runtime advisory, the version it would roll to, or why it cannot (no fixed version, fixed version too new, a runtime dependent rejects it, or a major bump of a package pinned in package.json).
  3. Apply: node .claude/skills/playwright-audit-deps/audit-deps.mjs --apply --report=<scratchpad>/deps-report.md. Pass --allow-major only when the user asked for a major bump; otherwise leave those packages in the "Not rolled" list and tell the user. Options: --max-age-days=N (default 8), --severity=low|moderate|high|critical (default moderate), --base=<ref> for the report diff.
  4. Verify: npm run build, then npm run itest tests/installation/bundle-licenses.spec.ts tests/installation/bundle-size.spec.ts and npm run check-dependency-age. A new dependency without a repository field, a bigger bundle, or a version younger than 7 days fails here. Fix the cause or drop that roll, then rerun the report.
  5. Commit package.json and package-lock.json as one commit:

    chore(deps): bump runtime deps
    
    <the markdown table and the "Not rolled" list from the report>
    
    Fixes: <issue url, when given>
  6. Report the table and the not-rolled packages to the user. Never push. When told to open the PR, use the same title and a body of one bullet naming the packages rolled for an advisory, the table, and the Fixes line.

Notes

  • npm audit lists a package that only depends on a vulnerable one (chromium-bidi via zod) as its own finding; the script follows the actual advisory, so such entries are not acted on.
  • A root-pinned package is bumped with npm install <name>@<version> --save-exact. Any other package is rewritten in place in package-lock.json and reconciled with npm install. The script fails if npm ls reports a problem afterwards.
  • Run the apply step once from the clean base; a second run labels the already rolled packages as plain dependencies in the "Reason" column, because the audit no longer flags them.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.