CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

review-against-ai-patterns

创建、编辑或审查任何人类可见产物时使用:方向决定前按风险读取反模式图鉴标本并建立 ANTI-SLOP-CHECK.md,制作中任一图鉴组成形即停下重构并记账,交付前对完整产物做全局扫描与反模板渲染审查;不代替领域 Skill 的专业门槛,也不裁决审美阶段。

69

Quality

87%

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

SKILL.md
Quality
Evals
Security

Quality

Content

92%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 tightly written procedural skill with excellent structure: concrete stop signals, explicit feedback loops, and clean one-level-deep reference delegation to files that all exist. The only trims are the migration-history sentence and related bookkeeping, which cost tokens without aiding execution.

Suggestions

Drop or compress the migration provenance sentence ('从 $visual-artifact-design 与 $build-polished-visual-frontends 无损迁出;领域 Skill 保留各自的停下信号与返工测试') to a single boundary clause — the migration history is not needed to execute the skill.

DimensionReasoningScore

Conciseness

The body is lean and assumes Claude's competence — every section is procedural with no concept explanations — but the migration-history sentence ('从 $visual-artifact-design 与 $build-polished-visual-frontends 无损迁出') is trimmable bookkeeping that does not help execute the skill, matching 'minor instances of over-explanation that could be trimmed'.

4 / 5

Actionability

Guidance is fully executable for an instruction-only skill: exact first file to read ('references/anti-slop-gallery/page-1.jpg 的索引'), a 2–4 page read budget, exact ledger location rules (cwd-relative ANTI-SLOP-CHECK.md), ten named stop-signal groups (G 灰色小字 through D 插画画风), and explicit reviewer-input rules plus a SELF_REVIEW_ONLY marker.

5 / 5

Workflow Clarity

Three clearly sequenced phases (before direction, during production, check & delivery) each carry explicit validation checkpoints and feedback loops: stop-and-refactor on any group forming, ledger 复核 required after modification and re-render, a mandatory pre-delivery global scan, and concrete completion conditions ('遗漏实际命中的组即未完成').

5 / 5

Progressive Disclosure

The ~45-line body is a pure overview that delegates details via clearly signaled, one-level-deep markdown links to real files: [账本格式](references/ledger-format.md), [反模板渲染审查](references/anti-slop-review.md), and the 14-page gallery directory — all verified to exist, with content appropriately split and easy to navigate.

5 / 5

Total

19

/

20

Passed

Description

81%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 dense, well-structured description that clearly states what the skill does across all three workflow phases and when to use it, with an explicit non-overlap boundary. Its main weakness is trigger vocabulary: it lacks the natural synonyms (设计, UI, 界面, 视觉, 页面) users would actually say when needing this skill.

Suggestions

Add natural trigger synonyms for the artifact category, e.g. 设计、UI、界面、页面、图表、海报, so users' everyday phrasing matches the trigger clause.

Consider naming one or two concrete deliverable examples (网页、海报、仪表盘) in the when-clause to sharpen matching without adding length.

Keep the existing explicit boundary statement ('不代替领域 Skill 的专业门槛,也不裁决审美阶段') — it is the strongest distinctiveness signal in the description.

DimensionReasoningScore

Specificity

The description lists multiple concrete actions spanning the full workflow — '按风险读取反模式图鉴标本并建立 ANTI-SLOP-CHECK.md', '任一图鉴组成形即停下重构并记账', '交付前…全局扫描与反模板渲染审查' — with a named artifact file, giving comprehensive coverage rather than the minor gaps of a 4.

5 / 5

Completeness

It explicitly answers both: what (read gallery by risk, build ANTI-SLOP-CHECK.md, stop-and-refactor on group formation, pre-delivery global scan and anti-template review) and when ('创建、编辑或审查任何人类可见产物时使用' — an explicit use-when clause with concrete trigger phrases).

5 / 5

Trigger Term Quality

Task verbs (创建、编辑、审查) and the trigger category '人类可见产物' are present, but common natural synonyms users would say — 设计、UI、界面、页面、视觉、图表、前端 — are missing, matching 'some relevant keywords but missing common variations or synonyms'.

3 / 5

Distinctiveness Conflict Risk

The niche (anti-AI-slop review) is distinct and explicitly bounded ('不代替领域 Skill 的专业门槛,也不裁决审美阶段'), but the very broad trigger '任何人类可见产物' creates minor overlap risk with closely related visual-design skills, so it is mostly distinct rather than minimal-risk.

4 / 5

Total

17

/

20

Passed

Validation

81%

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

CriteriaDescriptionResult

metadata_version

'metadata.version' is missing

Warning

metadata_field

'metadata' should map string keys to string values

Warning

referenced_paths_exist

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

Warning

Total

13

/

16

Passed

Repository

EverMind-AI/Raven

Path

plugins-dist/design-engine/raven_design/skills/review-against-ai-patterns/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.