chrome-devtools
Browser automation with Puppeteer CLI scripts. Use for screenshots, performance analysis, network monitoring, web scraping, form automation, or encountering JavaScript debugging, browser automation errors.
81
84%
Does it follow best practices?
Impact
57%
2.00xAverage score across 3 eval scenarios
Risky
Do not use without reviewing
Quality
Discovery
92%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong description that clearly identifies the tool (Puppeteer), lists concrete capabilities, and provides explicit trigger guidance with the 'Use for...' clause. The main weakness is that the scope is broad enough that some trigger terms (e.g., 'JavaScript debugging', 'performance analysis') could overlap with other skills. The phrasing 'or encountering JavaScript debugging, browser automation errors' is slightly awkward grammatically.
Suggestions
Tighten the phrasing around 'encountering JavaScript debugging, browser automation errors' — consider 'or when troubleshooting browser automation errors' for clarity.
Consider narrowing overlap-prone terms by qualifying them, e.g., 'browser-based performance analysis' or 'in-browser network monitoring' to reduce conflict with general debugging or performance skills.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: screenshots, performance analysis, network monitoring, web scraping, form automation, and JavaScript debugging/browser automation errors. | 3 / 3 |
Completeness | Clearly answers both what ('Browser automation with Puppeteer CLI scripts') and when ('Use for screenshots, performance analysis, network monitoring, web scraping, form automation, or encountering JavaScript debugging, browser automation errors'). | 3 / 3 |
Trigger Term Quality | Includes strong natural keywords users would say: 'screenshots', 'web scraping', 'form automation', 'browser automation', 'Puppeteer', 'network monitoring', 'performance analysis', 'JavaScript debugging'. These cover a good range of terms users would naturally use. | 3 / 3 |
Distinctiveness Conflict Risk | While 'Puppeteer' and 'browser automation' are fairly distinct, terms like 'web scraping', 'JavaScript debugging', and 'performance analysis' could overlap with general web development, scraping, or debugging skills. The scope is somewhat broad which increases conflict risk. | 2 / 3 |
Total | 11 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured, highly actionable skill with clear workflows and validation checkpoints. Its main weakness is moderate verbosity — the ImageMagick compression details, multi-OS installation instructions, and repeated pwd warnings inflate the token count without proportional value. The progressive disclosure could be improved by moving installation and detailed compression behavior to referenced files.
Suggestions
Move the detailed installation instructions (multi-OS support, ImageMagick setup) to a separate INSTALL.md and reference it from the Quick Start section to reduce token usage.
Condense the compression behavior section — the format-specific details (PNG resize percentages, JPEG quality levels) are implementation details Claude doesn't need; a brief note that auto-compression happens via ImageMagick suffices.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The content is mostly efficient but includes some unnecessary verbosity. The ImageMagick compression section is overly detailed with multiple format-specific behaviors that Claude could infer. The repeated 'CRITICAL: Always check pwd' warnings and the installation steps for multiple OS variants add bulk. However, most sections are reasonably lean. | 2 / 3 |
Actionability | Excellent actionability throughout — every script has concrete, executable command-line examples with real flags and expected JSON output. The chaining pattern, jq parsing, error recovery with snapshot.js fallback, and XPath alternatives are all copy-paste ready and specific. | 3 / 3 |
Workflow Clarity | Multi-step workflows are clearly sequenced with explicit validation checkpoints. The Execution Protocol section includes working directory verification before execution, output validation after operations (file existence check, JSON success verification, visual inspection), and a clear error recovery feedback loop (fail → snapshot → retry with different selector). | 3 / 3 |
Progressive Disclosure | The skill references `./scripts/README.md` for detailed script usage, which is good progressive disclosure. However, the compression behavior details, full installation instructions for multiple OSes, and troubleshooting could be split into separate files. The content is somewhat monolithic at ~170 lines when some sections (installation, compression details) could be referenced externally. | 2 / 3 |
Total | 10 / 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.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- secondsky/claude-skills
- Commit
5e92b71
- 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.