name: unreal-engine description: Unreal Engine single-entry skill with Hardness (H/M/L) and subagent orchestration for discovery, implementation, and verification. source:

• https://github.com/DSTN2000/claude-unreal-engine-skill
• https://github.com/sickn33/antigravity-awesome-skills/tree/main/skills/unreal-engine-cpp-pro
• https://github.com/404kidwiz/claude-supercode-skills/tree/main/game-developer-skill synced_at: 2026-03-31

Unreal Engine Skill — 入口索引

版本: 2.0 (原子化拆分版) 定位: 本文件是 导航索引 — 不包含实质规则定义，仅提供索引和查找 权威源映射: 每个章节下方标注了完整内容的存放位置

Purpose

Provide a single Unreal Engine skill entrypoint with deterministic Hardness governance, auditable delegation, and strict verification gates.

When To Use

• Any request involving Unreal Engine, UE5, .uproject, C++ gameplay systems, Blueprint + C++ integration, Enhanced Input, GAS, replication, or plugin behavior.

Non-Negotiables

• Zero assumptions: never invent asset names, module names, plugin states, or engine version.
• Discovery first, then implementation.
• Writer/reviewer separation for M/H.
• M/H requests are incomplete unless verification evidence is present.
• If reviewer separation is impossible, output Status: BLOCKED (never self-approve).
• Evaluate scriptability first; if a task is scriptable, prefer script/harness execution over manual repeated steps.
• Agent orchestration must strictly follow AgentHardness philosophy (intensity, delegation depth, and evidence depth must match risk/complexity).

核心框架索引

执行层索引

契约索引

工作流索引

Playbooks (详细执行指南)

References (按需加载)

Agent Prompt Files (子 Agent 初始化)

快速参考卡

Hardness 选择

后端选择

输出契约 (必需字段)

Hardness: L|M|H
Delegation Plan: [团队/角色/车道]
Scriptability: scriptable|non-scriptable + [path]
Execution Backend: mcp|cli-anything|native-cli
Verification Evidence: [通过/未验证项]
Status: DONE|BLOCKED

文档结构图

.trae/skills/unreal-engine/
├── SKILL.md                    # 本文件 (入口索引)
├── SOURCES.md                  # 来源声明
│
├── frameworks/                 # 核心框架
│   ├── hardness.md             # Hardness 分级规则
│   ├── delegation.md           # 委托与团队生成
│   └── scriptability.md        # 脚本化决策
│
├── execution/                  # 执行层
│   ├── backend_priority.md     # 后端优先级
│   └── cli_tool_layer.md       # CLI 工具层
│
├── contracts/                  # 契约定义
│   └── output_contract.md      # 输出契约
│
├── workflows/                  # 工作流
│   ├── execution_flow.md       # 执行流程
│   └── discovery.md            # 发现工作流
│
├── playbooks/                  # 详细执行指南
│   ├── discovery.md
│   ├── cpp_architecture.md
│   ├── multiplayer_and_net.md
│   ├── mcp_agent_orchestration.md
│   └── validation.md
│
└── references/                 # 参考文档 (按需加载)
    ├── mcp_tools_mab.md        # MAB 工具集
    ├── mcp_tools_umg.md        # UMG 工具集
    ├── mcp_tools_overlap.md    # 重叠分析
    ├── mcp_agent_system.md     # Agent 系统索引
    ├── mcp_extension_guide.md
    ├── enhanced_input.md
    ├── gameplay_ability_system.md
    ├── common_pitfalls.md
    └── cli_validation_report.md

文档版本: 2.0 (原子化拆分版) 最后更新: 2026-04-02 维护者: SoloCoder (AI Agent Orchestrator) 定位: 入口索引 — 本文件是导航中心，所有实质内容均有唯一权威源