webwright

Solve a user-specified web task code-as-action style by driving a local Playwright browser through one bash command at a time, saving screenshots and an action log into `final_runs/run_<id>/`, and visually verifying the result. Use when the user asks to automate a web task (search, filter, form-fill, multi-step flow, data extraction) and wants reusable scripts plus screenshot evidence rather than a one-shot answer.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

81%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, actionable skill body with a clear validated workflow and good progressive-disclosure organization. The main improvements are condensing the origin backstory and inlining a minimal executable browser-launch skeleton so the body is self-sufficient.

Suggestions

Condense the Webwright-origin narrative (lines 9-21) to a 2-3 line note; Claude does not need the JSON-wrapping/harness rationale to execute the skill.

Inline a minimal copy-paste Playwright launch+viewport+screenshot skeleton in the Explore or Workspace Contract section so the body is executable without opening the reference file.

Confirm the referenced `reference/playwright_patterns.md`, `reference/workflow.md`, and `reference/cli_tool_mode.md` are shipped alongside SKILL.md so the one-level-deep references resolve.

Dimension	Reasoning	Score
Conciseness	The body is mostly lean and concrete (workspace contract, workflow, hard rules), but the Webwright-origin backstory explaining the JSON-wrapping constraint and harness adaptation (lines 9-21) is rationale Claude does not need and could be trimmed to 2-3 lines.	4 / 5
Actionability	Provides concrete commands (`playwright install firefox`), exact config values (`viewport={"width": 1280, "height": 1800}`), file paths, a log format, and a plan.md template, but a full copy-paste browser-launch skeleton is deferred to the reference file rather than shown inline.	4 / 5
Workflow Clarity	A clear six-step sequence (Plan → Explore → Author → Execute → Self-verify → Done) with an explicit validation step, a fix→re-run→re-verify feedback loop, and a plan.md critical-points checklist for completion.	5 / 5
Progressive Disclosure	Well-organized sections with a dedicated Reference Files list pointing one level deep to playwright_patterns.md, workflow.md, and cli_tool_mode.md, each described; however the referenced reference/*.md files are not present in the review bundle to verify they ship as described.	4 / 5
	Total	17 / 20 Passed

Description

92%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description that clearly answers both what the skill does and when to use it, with concrete actions and low conflict risk. The only minor gap is missing a few common natural synonyms (scrape, crawl) in the trigger terms.

Dimension	Reasoning	Score
Specificity	Names multiple concrete actions — 'driving a local Playwright browser through one bash command at a time', 'saving screenshots and an action log into `final_runs/run_<id>/`', 'visually verifying the result' — giving comprehensive coverage of what the skill does.	5 / 5
Completeness	Explicitly states the 'what' (drive browser, save screenshots/log, visually verify) and an explicit 'Use when the user asks to automate a web task...' clause with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes natural user phrases like 'automate a web task (search, filter, form-fill, multi-step flow, data extraction)', but common variations such as 'scrape'/'crawl'/'web scraping' are absent, so coverage is good rather than comprehensive.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (code-as-action Playwright automation with screenshot evidence and run folders) with distinct triggers and a discriminator ('rather than a one-shot answer'), minimizing overlap with other skills.	5 / 5
	Total	19 / 20 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning

	Total	15 / 16 Passed

Repository: microsoft/Webwright
Path: skills/webwright/SKILL.md
Commit: bc26750

Reviewed: 29 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.