browser

用异步 Python 调用 QwenPaw 内置 Browser SDK 驱动真实浏览器。完整参考在下方；上下文压缩后请重新加载此 browser skill。

Quality

67%

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

Fix and improve this skill with Tessl

tessl review fix ./src/qwenpaw/agents/skills/browser-zh/SKILL.md

Quality

Content

81%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 and workflow-clear, with executable code and an explicit perceive-act-verify loop including validation checkpoints. Its main weakness is structure: the full API reference is inlined monolithically rather than split into referenced bundle files.

Suggestions

Move the large generated API reference into a separate file (e.g., references/browser-manual.md) and keep SKILL.md as a concise overview with a clear one-level-deep pointer to it.

If the SDK reference is auto-generated, keep the 'BEGIN/END GENERATED' block but relocate it to a bundle file referenced from the overview.

Add clear section headers (Quick start / Locator strategy / Human-only steps) so the inlined content is navigable even before any file split.

Dimension	Reasoning	Score
Conciseness	The body is largely lean and assumes Claude's competence (a copy-this-shape block plus a tight API table), with only minor redundant prose (the reload instruction restated in both the opening and the generated block), matching 'efficient; minor instances of over-explanation that could be trimmed.'	4 / 5
Actionability	It provides a fully executable copy-paste async Python example plus a complete, concrete API surface (signatures, return fields, locator methods), covering the common perceive→act→verify cases, matching 'fully executable; copy-paste ready code or commands; specific examples cover the common cases.'	5 / 5
Workflow Clarity	The perceive → act → verify loop is explicit with re-perceive checkpoints, a strict-mode check via count(), a step-down locator escalation, and a handoff-and-stop rule for human-only steps, matching 'clear sequence with explicit validation steps; feedback loops.'	5 / 5
Progressive Disclosure	The entire API reference is inlined as one monolithic generated block inside SKILL.md with no bundle files (references/, scripts/, assets/ all absent), so 'content that clearly belongs in separate files is inlined,' fitting the score-2 anchor despite some in-block structure.	2 / 5
	Total	16 / 20 Passed

Description

53%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 states a clear, specific purpose but omits any explicit trigger/when-to-use guidance and enumerates few concrete actions. It is distinct within the QwenPaw niche yet generic enough that it would not reliably surface on natural user phrasing.

Suggestions

Add an explicit 'Use when ...' clause listing natural user triggers (e.g., 'when the user asks to open, scrape, fill, or navigate a web page').

Enumerate several concrete actions the skill performs (open pages, read page text, click/fill elements, take screenshots) rather than a single high-level framing.

Include common synonyms and file/URL-style trigger words (web page, URL, form, click, screenshot) to broaden natural keyword coverage.

Dimension	Reasoning	Score
Specificity	It names the domain (browser automation) and a concrete framing ('驱动真实浏览器' via '异步 Python' + 'QwenPaw 内置 Browser SDK'), but does not enumerate multiple distinct actions, so it matches 'names domain and 1-2 concrete actions, but not comprehensive.'	3 / 5
Completeness	It gives a clear 'what' (drive a real browser via async Python + the Browser SDK) but no explicit 'when to use' clause; per the guidelines, a missing 'Use when...' caps this at 3.	3 / 5
Trigger Term Quality	It includes relevant terms ('browser', '真实浏览器', 'Browser SDK') but lacks the natural user-side phrasing and common synonyms/extensions users would say, fitting 'some relevant keywords but missing common variations.'	3 / 5
Distinctiveness Conflict Risk	The 'QwenPaw 内置 Browser SDK' framing carves a clear niche distinct from generic skills, with only minor overlap risk against other automation skills, matching 'mostly distinct; minor overlap risk.'	4 / 5
	Total	13 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
metadata_version	'metadata.version' is missing	Warning
metadata_field	'metadata' should map string keys to string values	Warning

	Total	14 / 16 Passed

Repository: agentscope-ai/QwenPaw
Path: src/qwenpaw/agents/skills/browser-zh/SKILL.md
Commit: f6dbc1d

Reviewed: 2 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.