Token导航 LogoToken导航

Agent Skill · 研究检索

level-design(关卡设计)

用于辅助界面设计、视觉规范、排版、配色、布局和交互体验优化。它适合让 Agent 根据产品场景整理页面结构、生成 UI 方案、检查视觉一致性或改进组件层级。使用时需要结合现有品牌、设计系统和用户任务,不应只堆装饰元素;涉及真实页面改动时,应通过截图或浏览器预览检查文本溢出、对齐和响应式表现。

level-design 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为只读;许可证为 MIT,暂无第三方扫描结果,按安装量在研究检索类的 73,675 个 Skill 里排第 72,863。

总安装 186周安装 8GitHub Stars 公开资料未说明下载 65核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
MIT

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:level-design(关卡设计)
来源仓库:https://github.com/jarrodmedrano/jarrod-claude-skills
仓库路径:skills/level-design
安装命令:
npx skills add jarrodmedrano/jarrod-claude-skills --skill "level-design"
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

AgentSkills.tonpx skills
npx skills add jarrodmedrano/jarrod-claude-skills --skill "level-design"

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

level-design 用于辅助界面设计、视觉规范和交互体验优化,支持 UI 方案生成。适用于 Codex、Claude、Cursor、Gemini CLI 等平台中的产品设计辅助场景。

通过 npx skills add jarrodmedrano/jarrod-claude-skills --skill "level-design" 命令安装。使用前需结合现有设计系统和用户任务,避免过度装饰影响可用性。建议通过浏览器预览验证文本溢出和对齐等实际效果。

来源仓库内路径:skills/level-design

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add jarrodmedrano/jarrod-claude-skills --skill "level-design";命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    它标注为只读,先用只读任务试一次,确认行为符合预期再扩大使用范围。

常见问题

6 条

level-design 怎么安装?

在终端执行 npx skills add jarrodmedrano/jarrod-claude-skills --skill "level-design"。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

level-design 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

level-design 需要什么权限?

权限级别是只读。该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

level-design 是免费的吗?

来源标注的许可证是 MIT。本站不代收任何费用,也不托管安装包;能否商用、要不要署名,以许可证原文为准。

level-design 用的人多吗?

公开数据是累计安装 186 次,近一周安装 8 次,下载 65 次。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

level-design 安全吗?

没有查到第三方扫描结果,安全性无法从公开资料判断,安装前请自行审查源码、依赖和命令行为。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 许可证明确为 MIT,商用和分发边界好判断。
  • 标注为只读权限,试用阶段不容易误改本地文件。

先想清楚的地方

  • 没有第三方扫描结果,安全性只能靠自己读源码判断。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。
  • 只想先试试效果、不希望它改动本地文件的人。

不太适合

  • 要求有安全扫描结论才肯安装的团队。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
level-design

关卡设计 · 本页

186—只读MIT未收录

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

Level Design Skill

A design consulting framework based on "Level Design: Processes and Experiences" edited by Christopher W. Totten (CRC Press, 2017).

Core Philosophy

Level design is the thoughtful execution of gameplay into gamespace for players to dwell in. It sits at the intersection of programming, design, and art—implementing the game design vision while leading players through experiences without revealing the designer's presence.

"Level designers don't merely create things for players to do. They create situations that invite players to interpret who they are." — Brian Upton

The Designer's Core Tasks

  1. Guide without forcing — Lead players through intended experiences while maintaining illusion of freedom
  2. Teach through space — Use environment to communicate mechanics, not tutorials
  3. Control pacing — Modulate intensity through spatial rhythm and stillness
  4. Support narrative — Align levels within overall game progression
  5. Create consistency — Establish and honor environmental rules players can rely on

Quick Reference: Level Types

TypeKey Considerations
LinearHide linearity through visual choice, narrative lures, environmental storytelling
Open-WorldPOI density, anchor locations, subregions, orientation landmarks
HorrorAnticipatory play, corners, one-way doors, visible-but-blocked escape
ProceduralHorizontal vs vertical integration of handmade content
Indie/FocusedExpand single core mechanic through level variation

Hiding Linearity

Players must feel in control even when following a predetermined path.

Techniques

TechniqueDescription
Coerced ProgressionTime pressure, pursuing enemies—no time to question the path
Environmental SignageIn-world signs, color coding (Mirror's Edge red)
NPC GuidesCompanions who lead, escort targets who follow, enemies to chase
Narrative LuresVisible objectives, story hooks that pull forward
Forced Choice IllusionBlock one path as player approaches, making "choice" feel organic

What Breaks the Illusion

  • Arbitrary locked doors
  • Invisible walls
  • Flimsy barriers (yellow police tape blocking a superhero)
  • Clear artificial constraints without narrative justification

See: references/hiding-linearity.md


Anticipatory Play & Horror Design

Horror isn't about jump scares—it's about dread before the corner.

The P.T. Framework

  • Corners — Hide what's ahead; players imagine horrors worse than you could show
  • Ratchet Doors — One-way progress; can't retreat, must face what's ahead
  • Valve Doors — Block progress temporarily; visible state reduces uncertainty about *if* blocked
  • Visible Escape — Show impossible exits to amplify feeling of being trapped

Key Principle

"Anticipatory play requires variety—the situation must evolve so players continually reassess. Static horrors become played out."

See: references/anticipatory-play.md


Open-World Planning (Burgess Method)

Three living documents for large-scale world design:

1. The World Map

  • Establish setting, scale, subregions
  • Plot anchor locations (story-critical, landmarks)
  • Include orienting features (visible from anywhere)
  • Plan natural boundaries (water, cliffs) over artificial walls

2. The Master List (Excel)

  • Every location with X/Y coordinates
  • Columns for: designer, quest associations, footprint size, difficulty, encounter type
  • Scatter graph overlay on map image
  • Filter to visualize distribution patterns

3. The Directory (Wiki)

  • Per-location pages with: status, goals, walkthrough, known issues, to-do
  • Category tags for filtering (by designer, by pass, by type)
  • Living documentation updated throughout development

POI Density

The frequency of points of interest defines exploration feel:

  • High density = Theme park feel (GTA cities)
  • Low density = Vast, sparse exploration (Just Cause countryside)
  • Non-uniform = Urban cores dense, rural areas sparse (Fallout 4)

See: references/open-world-planning.md


Play-Personas

Model player behavior before and after implementation.

The Process

  1. Analyze mechanics → Derive high-level behaviors from low-level actions
  2. Create matrix → Plot all behavioral combinations (2^n personas)
  3. Select cast → Choose 2-3 personas aligned with design vision
  4. Associate affordances → Link behaviors to spatial/ludic elements
  5. Orchestrate — Modulate which personas are viable throughout level
  6. Validate — Use telemetry to confirm players match expected personas

Example: Pac-Man

High-level behaviors: Center vs periphery dwelling, early vs late pill eating, linear vs broken paths

→ 8 persona combinations including "Fraidy Cat" (periphery, early pills, linear) and "Risk Taker" (center, late pills, broken)

See: references/play-personas.md


Themed Level Tropes

Classic environmental themes with established mechanics and expectations:

TropeCore ElementsDesign Advantages
Fire/IceEnvironmental hazards, timing puzzlesColor variety, physics tweaks
Dungeon/CavernTileable textures, traps, treasureEasily repeatable, expected danger
FactoryMoving platforms, conveyers, gearsRepurposable mechanics, scalable difficulty
JungleVines, branches, water, wildlifeFluid movement, colorful outdoor
SpookyAtmosphere, surprise, undeadCombines with any theme
PirateShips, treasure, melee, waterAction-ready, clear rewards
UrbanVerticality, cover, vehiclesReal-world familiarity
Space StationTech hazards, airlocks, zero-GSci-fi dungeon equivalent
SewerPipes, rats, rising waterModern dungeon stand-in

Mexican Pizza Technique: Combine two tropes for fresh results (fire + graveyard, jungle + urban ruins)

See: references/themed-environments.md


Indie Level Design Practices

When working with limited resources:

PracticeDescription
Expand Core MechanicOne strong mechanic explored through level variation (VVVVVV)
Iterative Level DesignRapid prototyping, playtest early and often
Design Modes Not LevelsCreate systems that generate challenge (endless runners)
Embrace EmergenceSimple rules, complex interactions
Object-Oriented DesignModular elements that combine predictably

Qualities of Good Level Design

  • Maintain flow: challenge without anxiety or boredom
  • Balance freedom with constraints (illusion of control)
  • Enable mastery and emergent solutions
  • Balance risk and reward proportionally
  • Guide without being obvious

See: references/indie-practices.md


Procedural vs Handmade Integration

Two integration models:

Vertical Integration

Handmade thread runs through procedural content (FTL quest chains, Spelunky secrets)

Best for: Narrative, puzzle sequences, coherent story beats

Horizontal Integration

Procedural and handmade content interchangeable in same slot (Dungeon Crawl vaults, URR buildings)

Best for: Ensuring specific gameplay moments, quality floors

Key Decision

Should players see which is which?

  • Yes (DCSS): Visual variety, risk/reward clarity
  • No (URR): Quality perception, seamless experience

See: references/procedural-handmade.md


Level Evaluation Framework

When evaluating a level design:

  1. Player Guidance: Can players find their way without obvious signposting?
  2. Pacing: Does intensity modulate appropriately? Are there moments of stillness?
  3. Teaching: Does the space teach mechanics before testing them?
  4. Consistency: Do environmental rules remain predictable?
  5. Persona Fit: Does the level support intended play styles?
  6. Density: Is POI distribution appropriate for the experience?
  7. Linearity Illusion: Do players feel in control of their path?

Common Pitfalls

PitfallSymptomSolution
Obvious RailsPlayer comments on being "on rails"Add visual choice, narrative justification
Empty SpacePlayers comment on emptinessIncrease POI density or justify sparseness
Lost PlayersPlayers wander aimlesslyAdd orientation landmarks, environmental cues
Played-Out ScaresHorror stops being scaryKeep threats evolving, limit exposure time
Arbitrary BarriersPlayers frustrated by blocked pathsUse narrative-justified or natural boundaries
Tutorial OverloadPlayers skip to "real game"Teach through safe early gameplay

Key Mantras

  • "Hide the designer's hand." Players should feel they're discovering, not being led.
  • "Corners are always significant." Transitions between visible and hidden create anticipation.
  • "POI density defines feel." Sparse = vast exploration; dense = action-packed.
  • "Static threats become furniture." Evolve dangers or limit exposure.
  • "The illusion of choice is enough." Players interpret forced choices as agency.
  • "Mexican pizza your themes." Combine familiar tropes for fresh experiences.

平台分布

windsurf

29.85%
按下载量换算19

OpenCode

23.42%
按下载量换算15

Codex

18.93%
按下载量换算12

Claude Code

14.14%
按下载量换算9

Antigravity

8.07%
按下载量换算5

Gemini CLI

3.67%
按下载量换算2

安全审计

暂无安全审计结果可展示。

权限和风险

只读

该 Skill 主要提供规则、说明或参考内容,本身偏只读;真正读写文件、联网或执行命令仍取决于宿主 Agent 的任务。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills