browser

Drive a live browser with async Python against QwenPaw's builtin Browser SDK. The full reference is below; re-load this browser skill after context compaction.

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-en/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 content is highly actionable with a clear perceive-act-verify workflow and executable code, but the full API reference is monolithically inlined in SKILL.md with no bundle files to offload it. Splitting the reference into a references/ file would improve progressive disclosure and conciseness.

Suggestions

Move the generated API reference block (BEGIN/END GENERATED: browser-manual) into a separate file under references/ and replace it with a concise pointer plus the most-used entrypoints.

De-duplicate the handoff/human-boundary guidance, which is restated in the intro, the handoff method entry, and the locator-fallback section.

Add a short Quick-start heading above the inline reference so the perceive-act-verify shape is visible before the dense API surface.

Dimension	Reasoning	Score
Conciseness	The body is mostly dense reference material with executable shape and minimal padding, though the large inlined API reference and repeated prose caveats (e.g., handoff guidance stated multiple times) could be trimmed slightly.	4 / 5
Actionability	It provides fully executable async Python covering connect/open/snapshot/act/verify with concrete method signatures, locator usage, and reading-result fields, giving copy-paste-ready guidance for the common cases.	5 / 5
Workflow Clarity	The perceive -> act -> verify loop is explicit, the locator fallback ladder is enumerated, and destructive/human-only steps are gated behind handoff with clear stop-and-verify checkpoints, including error-recovery guidance.	5 / 5
Progressive Disclosure	No bundle files exist (references/scripts/assets all absent) yet the entire ~160-line API reference is inlined in SKILL.md rather than split into separate files, matching the 'minimal structure; inlined content that belongs in a separate file' anchor.	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 clearly states what the skill does and distinguishes it from Playwright, but it lacks an explicit 'Use when...' trigger clause and enumerates few concrete browser actions. Adding natural trigger phrases and a couple of specific operations would lift specificity and completeness.

Suggestions

Add an explicit 'Use when...' clause naming natural triggers (e.g., 'Use when driving a live browser, controlling web pages, or automating web UIs').

List 2-3 concrete operations (e.g., navigate, fill forms, read page text, click elements) so the capability set is comprehensive.

Include common user synonyms or file/URL cues (e.g., 'web pages', 'sites', 'forms') to broaden trigger term coverage.

Dimension	Reasoning	Score
Specificity	Names the domain ('Drive a live browser') and one concrete action ('async Python against QwenPaw's builtin Browser SDK') but lists no specific browser operations like navigation, clicking, or text extraction, leaving coverage non-comprehensive.	3 / 5
Completeness	It clearly states the 'what' (drive a live browser with async Python via the SDK) but provides no explicit 'Use when...' clause, so the 'when' is only weakly implied; per the guidelines this caps completeness at 3.	3 / 5
Trigger Term Quality	It includes 'browser' as a relevant keyword but offers few natural user phrasings and no synonyms or variants beyond 'browser', so natural coverage is only partial.	3 / 5
Distinctiveness Conflict Risk	The phrase 'QwenPaw's builtin Browser SDK, not Playwright' carves out a clear niche and reduces conflict with general browsing/automation skills, with only minor overlap risk against generic web skills.	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-en/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.