arcanea-architect
Compare original and translation side by side
🇺🇸
Original
English🇨🇳
Translation
ChineseArcanea Architect Skill
Arcanea Architect Skill
Purpose
用途
This skill guides the design and implementation of the Arcanea ecosystem—the 64-agent creative intelligence system with multi-platform integration.
本Skill指导Arcanea生态系统的设计与实现——这是一个具备多平台集成能力的64-Agent创意智能系统。
When to Use
适用场景
- Designing agent architectures
- Implementing orchestration systems
- Creating multi-platform integrations
- Building creative AI tools
- Establishing systematic workflows
- 设计Agent架构
- 实现编排系统
- 创建多平台集成方案
- 构建创意AI工具
- 建立系统化工作流
Core Principles
核心原则
1. Excellence in Design
1. 卓越设计
- Every component must have clear purpose
- Architecture should be elegant, not clever
- User experience is paramount
- Code quality matters (tested, documented, maintainable)
- 每个组件必须具备明确用途
- 架构应简洁优雅,而非刻意炫技
- 用户体验是重中之重
- 重视代码质量(经过测试、文档完善、易于维护)
2. Attention to Detail
2. 注重细节
- No shortcuts in critical paths
- Validate all assumptions
- Document reasoning for decisions
- Test everything
- 关键路径绝不走捷径
- 验证所有假设
- 记录决策的推理过程
- 全面测试所有内容
3. Systematic Implementation
3. 系统化实施
- Build foundation first, features second
- Integration over isolation
- Progressive enhancement
- Backward compatibility when possible
- 先搭建基础,再开发功能
- 优先集成,避免孤立
- 渐进式增强功能
- 尽可能保持向后兼容性
Architecture Patterns
架构模式
Agent Registry Pattern
Agent注册模式
typescript
// Centralized agent definitions
interface Agent {
id: string; // Unique identifier
name: string; // Display name
court: string; // Elemental court
frequency: string; // Operating frequency
specialty: string; // What they do best
skills: string[]; // Available skills
prompts: Template[]; // Invocation templates
}typescript
// Centralized agent definitions
interface Agent {
id: string; // Unique identifier
name: string; // Display name
court: string; // Elemental court
frequency: string; // Operating frequency
specialty: string; // What they do best
skills: string[]; // Available skills
prompts: Template[]; // Invocation templates
}Conductor Pattern
指挥者模式
typescript
// Multi-agent orchestration
class Conductor {
async orchestrate(task: Task): Promise<Result> {
// 1. Analyze task complexity
// 2. Select optimal agent team
// 3. Determine execution strategy
// 4. Execute with monitoring
// 5. Learn from results
}
}typescript
// Multi-agent orchestration
class Conductor {
async orchestrate(task: Task): Promise<Result> {
// 1. Analyze task complexity
// 2. Select optimal agent team
// 3. Determine execution strategy
// 4. Execute with monitoring
// 5. Learn from results
}
}Skill Pattern
Skill模式
typescript
// Capability definitions
interface Skill {
id: string;
gate: number; // Which gate (1-10)
frequency: string; // Associated frequency
invocation: string; // How to trigger
process: Step[]; // Step-by-step procedure
agents: string[]; // Which agents can use
}typescript
// Capability definitions
interface Skill {
id: string;
gate: number; // Which gate (1-10)
frequency: string; // Associated frequency
invocation: string; // How to trigger
process: Step[]; // Step-by-step procedure
agents: string[]; // Which agents can use
}Implementation Checklist
实施检查清单
Phase 1: Foundation
第一阶段:基础搭建
- Agent registry (64 agents defined)
- Core conductor (orchestration engine)
- Skill system (50+ skills)
- AI router (opencode + Claude BYOK)
- Documentation architecture
- Agent注册中心(定义64个Agent)
- 核心指挥者(编排引擎)
- Skill系统(50+个Skill)
- AI路由(opencode + Claude BYOK)
- 文档架构
Phase 2: Integration
第二阶段:集成
- .claude/CLAUDE.md (Claude Code)
- .opencode/ skills (opencode editor)
- Tauri desktop app
- VS Code extension
- Obsidian plugin
- .claude/CLAUDE.md (Claude Code)
- .opencode/ skills (opencode编辑器)
- Tauri桌面应用
- VS Code扩展
- Obsidian插件
Phase 3: Polish
第三阶段:优化打磨
- UI/UX refinement
- Performance optimization
- Error handling
- User onboarding
- Testing & validation
- UI/UX优化
- 性能调优
- 错误处理
- 用户引导
- 测试与验证
Design Decisions
设计决策
Why 64 Agents?
为何选择64个Agent?
Decision: Use I Ching structure (8×8 = 64)
Reasoning:
- Complete symbolic system
- Sacred geometry (8 = cosmic order)
- Manageable yet comprehensive
- Each can spawn sub-agents
决策: 采用《易经》结构(8×8=64)
理由:
- 完整的符号系统
- 神圣几何(8代表宇宙秩序)
- 兼具可管理性与全面性
- 每个Agent可衍生子Agent
Why Ten Gates?
为何设置十个Gate?
Decision: Solfeggio frequency scale (174-1111Hz)
Reasoning:
- Historical resonance
- Metaphorical value
- Organizational clarity
- Memorable categorization
决策: 采用Solfeggio频率音阶(174-1111Hz)
理由:
- 历史共鸣性
- 隐喻价值
- 组织清晰度
- 易于记忆的分类方式
Why Hybrid AI?
为何采用混合AI方案?
Decision: opencode primary + Claude BYOK
Reasoning:
- Local-first (privacy, speed)
- Cost efficiency (free tier for most tasks)
- No vendor lock-in
- User control
决策: 以opencode为主 + Claude BYOK
理由:
- 本地优先(隐私性、速度)
- 成本效益(多数任务可使用免费层级)
- 无供应商锁定
- 用户可控
Why Tauri Desktop?
为何选择Tauri桌面框架?
Decision: Tauri over Electron
Reasoning:
- Smaller bundle (600KB vs 100MB)
- Rust performance
- Native OS integration
- Better security
决策: 选用Tauri而非Electron
理由:
- 包体积更小(600KB vs 100MB)
- Rust性能优势
- 原生系统集成
- 更高的安全性
Quality Standards
质量标准
Code Quality
代码质量
- TypeScript strict mode
- Comprehensive error handling
- Unit tests for core logic
- Integration tests for workflows
- Documentation for all public APIs
- TypeScript严格模式
- 全面的错误处理
- 核心逻辑单元测试
- 工作流集成测试
- 所有公开API均有文档
User Experience
用户体验
- Sub-500ms response times
- Clear error messages
- Progressive disclosure
- Offline capability
- Consistent cross-platform UI
- 响应时间低于500毫秒
- 清晰的错误提示
- 渐进式信息展示
- 离线使用能力
- 跨平台UI一致性
Performance
性能
- LRU caching for AI responses
- Lazy loading for agents
- Debounced UI updates
- Optimized renders
- Memory management
- AI响应结果的LRU缓存
- Agent懒加载
- UI更新防抖处理
- 渲染优化
- 内存管理
Resources
资源
Documentation
文档
- - Why 64 agents
AGENT_ARCHITECTURE_v4.md - - How it fits
IMPLEMENTATION_ARCHITECTURE.md - - AI integration
BYOK_SAAS_ARCHITECTURE.md - - Skills system
SKILL_ARCHITECTURE_ANALYSIS.md
- - 为何采用64个Agent
AGENT_ARCHITECTURE_v4.md - - 系统适配方式
IMPLEMENTATION_ARCHITECTURE.md - - AI集成方案
BYOK_SAAS_ARCHITECTURE.md - - Skill系统分析
SKILL_ARCHITECTURE_ANALYSIS.md
Code
代码
- - Agent registry and conductor
arcanea-agents/ - - Tauri desktop application
desktop/ - - Claude Code integration
.claude/ - - opencode editor integration
.opencode/
- - Agent注册中心与指挥者
arcanea-agents/ - - Tauri桌面应用
desktop/ - - Claude Code集成
.claude/ - - opencode编辑器集成
.opencode/
Templates
模板
- Use template for new skills
SKILL.md - Use agent template for new agents
- Use component templates for UI
- 新建Skill使用模板
SKILL.md - 新建Agent使用Agent模板
- UI组件使用组件模板
Anti-Patterns to Avoid
需避免的反模式
❌ Don't: Create agents without clear purpose
✅ Do: Every agent solves specific problems
❌ Don't: Use frequencies as acoustic prescriptions
✅ Do: Use as metaphorical categories
❌ Don't: Build features before foundation
✅ Do: Foundation first, features second
❌ Don't: Lock users into specific AI provider
✅ Do: Support multiple providers (BYOK)
❌ Don't: Over-engineer simple solutions
✅ Do: Elegant simplicity
❌ 切勿: 创建无明确用途的Agent
✅ 应当: 每个Agent解决特定问题
❌ 切勿: 将频率作为声学指令
✅ 应当: 将其用作隐喻分类
❌ 切勿: 在基础未完成前开发功能
✅ 应当: 先搭建基础,再开发功能
❌ 切勿: 将用户锁定在特定AI供应商
✅ 应当: 支持多供应商(BYOK)
❌ 切勿: 对简单解决方案过度设计
✅ 应当: 追求简洁优雅
Success Metrics
成功指标
- Technical: 100% test pass rate, <500ms response time
- User: Can complete tasks without documentation
- Adoption: Daily active users, retention rate
- Satisfaction: User feedback, feature requests
- 技术层面: 测试通过率100%,响应时间<500ms
- 用户层面: 无需文档即可完成任务
- 采用率: 日活跃用户数、留存率
- 满意度: 用户反馈、功能需求
Conclusion
结论
Building Arcanea requires:
- Vision - Clear understanding of what we're building
- Discipline - Following architecture, not shortcuts
- Craft - Attention to every detail
- Iteration - Build, test, refine, repeat
The foundation determines the height. Build it well.
This skill should be used whenever implementing or extending the Arcanea ecosystem.
构建Arcanea需要:
- 愿景 - 清晰理解我们要构建的系统
- 纪律 - 遵循架构规范,不走捷径
- 匠心 - 注重每一个细节
- 迭代 - 构建、测试、优化、重复
基础决定高度,务必扎实搭建。
在实施或扩展Arcanea生态系统时,应使用本Skill。