CtrlK
BlogDocsLog inGet started
Tessl Logo

browser-automation

统一自动化入口。覆盖浏览器自动化(Playwright)和 Windows 桌面应用自动化(OpenReverse)。 浏览器场景:打开网页、点击、填表、爬取、截图、自动化登录、渗透页面交互。 桌面场景:操作 IDA/x64dbg 等 GUI 工具、Windows UI Automation、视觉驱动交互、桌面应用网络抓包。 触发关键词:浏览器自动化、桌面自动化、打开网页、填表、爬取、截图、自动化登录、Playwright、agent-browser、headless、OpenReverse、UIA、CUA、桌面操作、Windows 自动化。

70

Quality

85%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Critical

Do not install without reviewing

SKILL.md
Quality
Evals
Security

Quality

Content

77%Weight 40%Scale 1-3

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

The content is highly actionable with executable commands, clear sequencing, validation checkpoints, and a completion checklist. Its weaknesses are redundancy (duplicated command reference and repeated install steps) and poor use of bundle files: the cheatsheet and bootstrap script are never referenced from the body.

Suggestions

Replace the inline '命令参考' block with a one-line pointer to references/playwright-cheatsheet.md to remove the duplication and actually signal the bundled reference.

Reference scripts/setup.ps1 by path in the bootstrap section (e.g., '缺 Playwright 时运行 scripts/setup.ps1') instead of only the word 'bootstrap', so the bundled script is discoverable.

Consolidate the OpenReverse install steps so they appear once, with the '手动安装引导' block linking back rather than repeating the full sequence.

DimensionReasoningScore

Conciseness

The body is mostly efficient executable commands and tables without concept padding, but the inline '命令参考' duplicates the bundled playwright-cheatsheet.md and the OpenReverse install steps appear twice ('安装与配置' and '手动安装引导'), so it could be tightened — fitting the 'mostly efficient but could be tightened' anchor rather than 'every token earns its place'.

2 / 3

Actionability

Provides fully executable, copy-paste-ready commands throughout — 'agent-browser open <url>', 'agent-browser snapshot -i', 'git clone ...', 'npm run init:agents -- --target=all', 'npm run doctor:network' — plus concrete decision tables, matching the 'fully executable code/commands; copy-paste ready' anchor.

3 / 3

Workflow Clarity

Sequences are explicit (ACTION REQUIRED NOW/NEXT/ACT, the 4-step browser 核心工作流) with explicit validation checkpoints ('agent-browser close' to prevent leaks, 'wait --load networkidle', 'npm run doctor:cua-runtime/doctor:network') and a final '任务完成自检' checklist, matching the anchor requiring clear sequence + explicit validation + checklists.

3 / 3

Progressive Disclosure

Bundle files references/playwright-cheatsheet.md and scripts/setup.ps1 exist but are never linked or signaled from the body, while the command reference that belongs in the cheatsheet is inlined inline — matching 'content that should be separate is inline' and 'references present but not clearly signaled' rather than the well-signaled one-level-deep anchor.

2 / 3

Total

10

/

12

Passed

Description

92%Weight 40%Scale 1-3

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

The description is strong: it enumerates concrete capabilities across both browser and desktop domains in third person and supplies an explicit trigger-keyword list for routing. Its main weakness is overlap risk from generic web terms shared with narrower browser skills.

Suggestions

Tighten distinctiveness by leading with the tool-anchored triggers (Playwright/agent-browser for browser, OpenReverse for Windows desktop) and demoting generic terms like '爬取/截图' to secondary mentions.

Consider an explicit 'Use when the target is a web page or a Windows desktop GUI app' clause to make the when-guidance structural rather than only keyword-based.

DimensionReasoningScore

Specificity

Lists many concrete actions across both domains — '打开网页、点击、填表、爬取、截图、自动化登录、渗透页面交互' and '操作 IDA/x64dbg 等 GUI 工具、Windows UI Automation、视觉驱动交互、桌面应用网络抓包' — matching the 'multiple specific concrete actions' anchor rather than the partial score-2 anchor.

3 / 3

Completeness

Clearly answers 'what' (browser automation via Playwright + desktop automation via OpenReverse with enumerated scenarios) and provides explicit 'when' guidance via the 触发关键词 list, which counts as equivalent explicit trigger guidance and avoids the missing-trigger cap of 2.

3 / 3

Trigger Term Quality

The explicit 触发关键词 list gives good coverage of natural user terms ('打开网页、填表、爬取、截图、自动化登录、Playwright、Windows 自动化'), matching the 'good coverage of natural terms' anchor; a few jargon entries (UIA, CUA, headless) do not undermine the overall natural coverage.

3 / 3

Distinctiveness Conflict Risk

Named-tool triggers (Playwright, agent-browser, OpenReverse, UIA, CUA) are distinctive, but the broad dual scope plus generic web terms ('爬取、截图、填表、打开网页') could still overlap with a dedicated web-scraping or browser skill, fitting the 'somewhat specific but could still overlap' anchor rather than a clear niche.

2 / 3

Total

11

/

12

Passed

Validation

100%

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

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
zhaoxuya520/reverse-skill
Reviewed

Table of Contents

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.