cdp

Drive Chrome via the DevTools Protocol from JavaScript. Run JS snippets through the `browser-harness-js` CLI — it auto-spawns a long-lived bun HTTP server holding a fully-typed CDP `Session`, and every call (`browser-harness-js 'await session.Page.navigate(...)'`) executes against the same persistent connection. Session, active target, and globals survive across calls. Use when the user wants to automate, script, or inspect a Chrome browser via CDP — single tab or multi-tab, attach to existing Chrome or to a new one launched with --remote-debugging-port.

Quality

95%

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.

A well-organized, highly actionable body with explicit workflows and validation loops for attaching to a running Chrome. The main weakness is progressive disclosure: the doc references an sdk/ bundle (generated.ts, repl.ts, etc.) that is not present in the reviewed skill directory.

Suggestions

Ship the referenced sdk/ bundle files (generated.ts, repl.ts, session.ts, gen.ts, protocol JSONs, the CLI) so the documented one-level-deep references actually resolve.

Tighten the Connecting section's prose (auto-detect rationale and per-OS profile path enumeration) by moving exhaustive OS path lists into a reference file, keeping the main path lean.

Add an explicit 'verify the SDK files exist at <skill-dir>/sdk' step in Setup so a missing bundle is caught before first use rather than at first method lookup.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient and assumes Claude's competence, but the prose around auto-detect, per-OS profile paths, and timeout rationale adds some explanation that could be trimmed; it is noticeably above the midpoint but not perfectly lean.	4 / 5
Actionability	Copy-paste-ready bash and JS snippets cover the common cases (connect, navigate, evaluate, pick a target, events, persisting state), with a clear results table and concrete error-detection patterns.	5 / 5
Workflow Clarity	The 'Connecting to a running Chrome' section gives an explicit 5-step sequence with validation feedback loops (auto-detect fails -> open inspect -> tick Discover -> Allow -> retry / bump timeoutMs), and setup includes a verify-the-CLI-is-reachable checkpoint.	5 / 5
Progressive Disclosure	Content is well-sectioned with one-level-deep references (sdk/generated.ts, sdk/repl.ts, sdk/session.ts) listed in a Files section, but the SDK bundle files are not actually present in the skill directory so the referenced structure cannot be fully navigated.	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.

A strong, third-person description that concretely names capabilities and gives explicit trigger guidance for the CDP-automation niche. It is dense but every clause earns its place, with minimal conflict risk.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions (drive Chrome via CDP, run JS snippets, spawn a persistent server, execute calls against a persistent connection) with comprehensive coverage of the tool's behavior and state persistence.	5 / 5
Completeness	Clearly states what (drive Chrome via CDP, run JS snippets through a CLI that holds a persistent typed Session) and when ('Use when the user wants to automate, script, or inspect a Chrome browser via CDP'), both with concrete triggers.	5 / 5
Trigger Term Quality	Covers natural user phrases ('automate, script, or inspect a Chrome browser via CDP', 'single tab or multi-tab', 'attach to existing Chrome') plus technical terms (CDP, --remote-debugging-port) with good synonym coverage.	5 / 5
Distinctiveness Conflict Risk	The CDP/DevTools Protocol niche with explicit triggers (single/multi-tab, attach-or-launch) is distinct and unlikely to fire for unrelated skills.	5 / 5
	Total	20 / 20 Passed

Validation

100%

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 — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: browser-use/browser-harness-js
Path: SKILL.md
Commit: 2d9a5ed

Reviewed: 14 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.