agent-browser

Browser automation for AI agents using the agent-browser CLI and Playwright. Use when navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, logging into sites, or automating any browser task. Triggers on "open a website", "fill out a form", "click a button", "scrape data", "test this web app", "automate browser actions", or any programmatic web interaction request.

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

88%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.

The body is highly actionable with a clear sequenced workflow, verification feedback loops, and well-organized one-level-deep references. The two weaknesses are time-sensitive version/date info outside a deprecated section and dangling references to a non-existent templates directory.

Suggestions

Move the Validation Status dates and version numbers (e.g. 2026-02-28, agent-browser 0.15.1, Chromium chromium-1194) into a clearly labeled 'Validation / version notes' or deprecated-style section so time-sensitive info does not penalize conciseness.

Create the templates/ directory with form-automation.sh, authenticated-session.sh, and capture-workflow.sh, or remove the Ready-to-Use Templates section to eliminate the dangling references.

Tighten a few explanatory passages (e.g. the Configuration File priority sentence) into tighter reference form to lift conciseness from 4 toward 5.

Dimension	Reasoning	Score
Conciseness	The body is a dense, command-focused reference that assumes Claude's intelligence (no padding about what browsers/Playwright are), but the Validation Status table embeds time-sensitive dates and version numbers (2026-02-28, agent-browser 0.15.1, Chromium chromium-1194) outside a deprecated/old-patterns section, which the guidelines explicitly penalize.	4 / 5
Actionability	Commands throughout are fully executable and copy-paste ready, with concrete worked examples covering the common cases (form submission, authentication, data extraction, diffing, eval).	5 / 5
Workflow Clarity	The core workflow is clearly sequenced (Navigate → Snapshot → Interact → Re-snapshot), the Ref Lifecycle section gives an explicit re-snapshot checklist, diff snapshot provides a verification feedback loop, and the Error Recovery section gives symptom→fix loops.	5 / 5
Progressive Disclosure	The Deep-Dive Documentation table signals one-level-deep references with a "When to Use" column and all 7 referenced files exist, but the Ready-to-Use Templates section references templates/*.sh files that do not exist (no templates directory), a broken-navigation gap that keeps it below 5.	4 / 5
	Total	18 / 20 Passed

Description

100%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.

The description is strong across all dimensions: it states concrete capabilities, provides comprehensive natural trigger phrases, and explicitly covers both what and when in third person. Minor breadth in "any browser task" does not materially raise conflict risk given the specific tool binding.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions (navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, logging into sites) — comprehensive coverage matching the score-5 anchor.	5 / 5
Completeness	Explicitly answers both what (browser automation via agent-browser CLI and Playwright) and when ("Use when..." plus a "Triggers on..." clause with concrete trigger phrases).	5 / 5
Trigger Term Quality	Natural user phrases plus synonyms are covered comprehensively: "open a website", "fill out a form", "click a button", "scrape data", "test this web app", "automate browser actions".	5 / 5
Distinctiveness Conflict Risk	Browser automation via a specific CLI/Playwright is a clear niche with distinct triggers and minimal conflict risk; the broad "any browser task" phrasing is a minor breadth over-claim, not enough to drop to 4.	5 / 5
	Total	20 / 20 Passed

Validation

81%

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 — 13 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (502 lines); consider splitting into references/ and linking	Warning
allowed_tools_field	'allowed-tools' contains unusual tool name(s)	Warning
relative_links	Relative link issues: 3 missing	Warning

	Total	13 / 16 Passed

Repository: Jamie-BitFlight/claude_skills
Path: .claude/skills/agent-browser/SKILL.md
Commit: 3e0b2af

Reviewed: 3 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.