browser
Web browser automation with AI-optimized snapshots for claude-flow agents
54
43%
Does it follow best practices?
Impact
57%
2.52xAverage score across 3 eval scenarios
Risky
Do not use without reviewing
Optimize this skill with Tessl
npx tessl skill review --optimize ./.claude/skills/browser/SKILL.mdQuality
Discovery
22%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 too vague and jargon-heavy, failing to list concrete actions or provide explicit trigger guidance. 'AI-optimized snapshots' is unclear buzzword language, and the mention of 'claude-flow agents' is implementation-specific rather than user-facing. It needs substantial improvement in specificity, trigger terms, and completeness.
Suggestions
List specific concrete actions such as 'navigate to URLs, click elements, fill forms, extract page content, take screenshots' to improve specificity.
Add an explicit 'Use when...' clause with natural trigger terms like 'open a webpage', 'scrape website', 'automate browser', 'interact with web page', 'browser testing'.
Replace jargon like 'AI-optimized snapshots' with plain language describing what the feature actually does, e.g., 'captures accessible page snapshots for analysis'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | The description says 'web browser automation' which is a broad domain reference, and 'AI-optimized snapshots' is vague jargon. No concrete actions are listed (e.g., clicking elements, filling forms, scraping data, navigating pages). | 1 / 3 |
Completeness | The 'what' is vaguely stated as 'web browser automation with AI-optimized snapshots' but lacks specifics. There is no 'when' clause or explicit trigger guidance at all, which per the rubric should cap completeness at 2, and the weak 'what' brings it to 1. | 1 / 3 |
Trigger Term Quality | 'Web browser automation' and 'snapshots' are somewhat relevant keywords, but missing common user terms like 'scrape', 'click', 'navigate', 'open webpage', 'fill form', 'screenshot', 'headless browser', 'Puppeteer', 'Playwright', or 'selenium'. | 2 / 3 |
Distinctiveness Conflict Risk | 'Web browser automation' narrows the domain somewhat, and 'claude-flow agents' adds some specificity to the context. However, it could overlap with general web scraping, testing, or other browser-related skills due to lack of concrete action differentiation. | 2 / 3 |
Total | 6 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid, actionable reference skill with excellent command documentation via tables and concrete examples. Its main weaknesses are the lack of error handling/validation guidance in workflows (what to do when interactions fail, how to verify page state) and the monolithic structure that could benefit from splitting detailed reference material into separate files. The content is mostly concise but includes some sections (Hooks, Memory Integration) that add bulk without strong justification.
Suggestions
Add error recovery guidance: what to do when element refs are stale, when clicks don't trigger expected navigation, or when waits time out — include a validate-and-retry pattern.
Split the detailed command reference tables and Claude Flow integration into separate referenced files (e.g., COMMANDS.md, INTEGRATION.md) to keep SKILL.md as a focused overview.
Remove or significantly trim the Hooks section, which feels tangential to browser automation and adds tokens without clear value.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient with good use of tables and code blocks, but includes some unnecessary sections (Hooks, Memory Integration) that feel like padding, and the tips section restates things already implied by the workflow. The intro line about '93% context reduction' is a nice touch but the overall document could be tightened. | 2 / 3 |
Actionability | Provides fully executable, copy-paste ready commands throughout. The core workflow, quick reference tables, selector examples, and multi-step examples (login flow, form submission, data extraction) are all concrete and specific with real command syntax. | 3 / 3 |
Workflow Clarity | The core workflow shows a clear 4-step sequence (navigate → snapshot → interact → re-snapshot), but there are no validation checkpoints or error recovery steps. What happens if a click fails? If an element ref is stale? The examples show linear happy-path flows without any feedback loops or error handling guidance. | 2 / 3 |
Progressive Disclosure | The content is well-organized with clear section headers and tables, but it's a monolithic document with no references to external files. The Claude Flow integration section and detailed command reference tables could be split into separate files. For a skill of this length (~150 lines), some progressive disclosure into supporting files would improve scannability. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- ruvnet/ruvector
- Commit
c2089c4
- 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.