CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

browser-preview

Hub 内嵌浏览器预览 localhost 应用。 Use when: 写前端代码、跑 dev server、需要看页面效果、调 UI、operator说"看看效果"。 Not for: 后端纯 API 开发、不涉及页面的工作。 Output: 前端页面在 Hub browser panel 中实时预览。

Quality

78%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide
SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./cat-cafe-skills/browser-preview/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

70%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.

An operationally rich body with excellent workflow gating and error-recovery tables, weakened by noticeable redundancy (the launcher command and grammar rules repeated across sections) and a monolithic single-file structure with no progressive disclosure despite its length.

Suggestions

Deduplicate the managed-launcher start command and consolidate the bounded-vs-ordinary `--dir`/`--cwd` grammar rules into one section; the current body restates them in four places (基础流程, Step 1, 常见错误, 什么时候主动用), which could be cut by roughly a third.

Move the 技术要点 table (Preview Gateway, iframe sandbox, delivery contract details) and the launcher grammar contract into a reference file (e.g. references/preview-architecture.md) and keep a one-line signaled pointer in SKILL.md, enabling progressive disclosure.

DimensionReasoningScore

Conciseness

The content is almost entirely non-obvious operational contract detail rather than padded explanation, but the managed-launcher start command is duplicated in two full blocks and the bounded-vs-ordinary `--dir`/`--cwd` grammar rule is restated in four different sections. Anchor 3 (could be tightened); not 4 because the repetition goes beyond minor trimming.

3 / 5

Actionability

Concrete, near-copy-paste commands ("pnpm --dir /absolute/cat-cafe-feature-worktree preview:process start --port PORT ... pnpm dev:direct", "curl -s -o /dev/null -w %{http_code} http://localhost:PORT") with full output interpretation and a deliveryStatus decision table. Not 5: Step 2's MCP call embeds prose placeholders ("当前 worktreeId(有就传)") and the `blocked` branch just says "看 deliveryReason" without enumerating reasons or fixes.

4 / 5

Workflow Clarity

Steps 1-3 are strictly sequenced with hard validation gates ("只有 status=running 才进入下步", "200/301/304 = 可以继续", "000 = 服务器没起来,先启动再说"), a deliveryStatus feedback loop before claiming success, and a phenomenon→cause→fix error-recovery table. Matches the anchor-5 pattern of validate-then-proceed with error recovery.

5 / 5

Progressive Disclosure

Good section structure, but there are no reference files at all: the ~145-line body inlines the tech-details table and launcher grammar contracts that belong in separate reference material, and the single external pointer ("../.cat-cafe-shared-refs/evidence-output-contract.md") is buried in a bullet rather than clearly signaled. Not 2 because the sections are well-organized; not 4 because content that should be split out is entirely inline.

3 / 5

Total

15

/

20

Passed

Description

78%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: explicit what/when/not-for boundaries with natural bilingual trigger phrases and a clear Hub-embedded-preview niche. Its main limitation is that it undersells the skill's actual capabilities, listing only the core preview action.

DimensionReasoningScore

Specificity

"Hub 内嵌浏览器预览 localhost 应用" and "Output: 前端页面在 Hub browser panel 中实时预览" name the domain with 1-2 concrete actions, but the description doesn't enumerate the fuller capability set (managed process lifecycle, screenshots, console capture) that the body demonstrates. Not 4: it does not list several specific actions.

3 / 5

Completeness

Explicitly answers what ("Hub 内嵌浏览器预览 localhost 应用"), when ("Use when: 写前端代码、跑 dev server、需要看页面效果、调 UI"), and adds "Not for" and "Output" boundaries with concrete trigger phrases. Matches the anchor-5 example structure exactly.

5 / 5

Trigger Term Quality

"写前端代码、跑 dev server、需要看页面效果、调 UI、看看效果" are natural phrases a user would actually say, in both Chinese and English. Not 5: common variations like "preview", "HMR", "浏览器预览" appear only in the separate triggers list, not the description itself.

4 / 5

Distinctiveness Conflict Risk

The "Hub 内嵌浏览器" framing carves a clear niche distinct from system-browser tools, and "Not for: 后端纯 API 开发、不涉及页面的工作" bounds it. Not 5: the "跑 dev server" trigger overlaps a generic dev-server skill, a minor conflict risk.

4 / 5

Total

16

/

20

Passed

Validation

87%

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

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

referenced_paths_exist

Referenced path issues: 1 missing, 1 deeper-than-1-level

Warning

Total

14

/

16

Passed

Repository

zts212653/clowder-ai

Path

cat-cafe-skills/browser-preview/SKILL.md

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.