CtrlK
BlogDocsLog inGet started
Tessl Logo

browser

Automate web browser interactions using natural language via CLI commands. Use when the user asks to browse websites, navigate web pages, extract data from websites, take screenshots, fill forms, click buttons, or interact with web applications.

68

Quality

83%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Low

Low-risk findings worth noting

The canonical home for this skill is browser in UnicomAI/wanwu

SKILL.md
Quality
Evals
Security

Quality

Content

78%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 lean, highly actionable body with copy-paste-ready commands, a sensible workflow with a verify step, and good section organization. Its main defect is that all three referenced files (EXAMPLES.md, REFERENCE.md, setup.json) are missing from the bundle, which undermines the progressive-disclosure structure and the setup instructions.

Suggestions

Add the referenced EXAMPLES.md and REFERENCE.md files to the skill bundle, or remove the dangling links at the end of SKILL.md so navigation does not break.

Ship the referenced setup.json (or replace the 'check setupComplete' instruction with a check that does not depend on a missing file) so the first-time setup flow is executable as written.

Trim the environment-selection bullets to a single if/else statement and drop low-value table rows (e.g., Speed) to tighten conciseness; consider consolidating Best Practices and Troubleshooting into one ordered workflow with an explicit act-verify-retry loop.

DimensionReasoningScore

Conciseness

The body is efficient with concrete command blocks and no concept explanations Claude already knows, but minor instances could be trimmed — e.g., the redundant bullet 'No user prompting: The selection happens automatically' and low-value table rows like 'Speed | Faster | Slightly slower'. This fits 'minor instances of over-explanation' rather than 'every token earns its place'.

4 / 5

Actionability

Fully executable, copy-paste-ready content: a complete command reference ('browser navigate <url>', 'browser act "<action>"'...), setup commands ('npm install', 'npm link'), and a Quick Example covering the common navigate/act/extract/close flow. It is not 4 because there are no gaps in executable guidance.

5 / 5

Workflow Clarity

A clear sequence with checkpoints exists — 'Always navigate first', 'View screenshots after each command to verify', 'Close browser when done', plus troubleshooting recovery ('Action fails: Use browser observe') — but the workflow is scattered across Commands, Best Practices, and Troubleshooting sections rather than one coherent sequence with an explicit validate-fix-retry loop, keeping it below the 5 anchor.

4 / 5

Progressive Disclosure

Sections are well-organized and references are clearly signaled ('For detailed examples, see [EXAMPLES.md](EXAMPLES.md)'), but EXAMPLES.md, REFERENCE.md, and setup.json are referenced and none exist in the bundle — following any of them breaks navigation. Scored against the actual (empty) bundle structure, this is a structural gap beyond the 'minor organization gaps' of the 4 anchor.

3 / 5

Total

16

/

20

Passed

Description

88%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 description that clearly states what the skill does and when to use it, with concrete actions and natural trigger phrases in third-person voice. The only weaknesses are a few missing common synonyms (e.g., web scraping) and slight overlap risk with screenshot/scraping skills.

DimensionReasoningScore

Specificity

Names the mechanism ('using natural language via CLI commands') and six concrete actions — 'navigate web pages, extract data from websites, take screenshots, fill forms, click buttons' — giving comprehensive, multi-action coverage matching the top anchor; it is not 4 because no meaningful capability gaps remain.

5 / 5

Completeness

Explicitly answers both questions: 'what' ('Automate web browser interactions using natural language via CLI commands') and 'when' ('Use when the user asks to browse websites...') with concrete trigger phrases; a missing 'Use when' clause would cap it at 3, but it is fully present.

5 / 5

Trigger Term Quality

Good natural-phrase coverage ('browse websites', 'take screenshots', 'fill forms', 'click buttons'), but common variations users would actually say such as 'web scraping', 'scrape', or 'automate' are absent, placing it just below the comprehensive-synonyms anchor.

4 / 5

Distinctiveness Conflict Risk

Browser automation is a mostly distinct niche, but 'take screenshots' and 'extract data from websites' create minor overlap risk with dedicated screenshot or scraping skills, so it does not reach the minimal-conflict anchor.

4 / 5

Total

18

/

20

Passed

Validation

93%

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

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 2 missing

Warning

Total

15

/

16

Passed

Repository
UnicomAI/wanwu
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.