规格驱动的多 Agent 协同工作流框架 — 融合 Superpowers 技能引擎与 Spec Kit 规格模板
npx claudepluginhub 30675137/superspec当需要汇总任务状态、输出过程指标、识别阻塞项或生成项目状态看板时派发此 Agent
当阶段 Skill 需要文档命名校验、模板合规检查、版本规范验证或输出结构检查时派发此 Agent
当阶段 Skill 需要按任务清单落地实现、产出文档/代码/配置、提交实现说明时派发此 Agent
当阶段 Skill 需要查询历史方案、已有规范、历史决策或提供复用建议时派发此 Agent
当需要判断项目当前状态、决定下一阶段、处理阶段流转或管理升级时派发此 Agent
当阶段 Skill 需要目标澄清、需求归纳、用户场景分析、价值判断或优先级排序时派发此 Agent
当阶段 Skill 需要任务拆解、里程碑设计、Sprint 规划、依赖识别或资源排期时派发此 Agent
当需要执行 spec/design/delivery review、质量门禁检查或验收评审时派发此 Agent
当项目完成需要复盘、偏差分析、知识沉淀或模板改进建议时派发此 Agent
当阶段 Skill 需要系统设计、模块划分、边界定义、风险评估或技术取舍时派发此 Agent
在 Deliver 阶段完成后、进入 Verify 阶段前触发，检查交付完整性
在 Design 阶段完成后、进入 Deliver 阶段前触发，审查 Design 文档质量
在 Define 阶段完成后、进入 Design 阶段前触发，审查 Spec 文档质量
Use when Discover is complete and confirmed, to formalize requirements into a structured specification
Use when Design has passed review, to break down into tasks and execute implementation
Use when Spec has passed review and technical design is needed before task breakdown
Use when starting a new project, receiving a vague idea, or needing business clarification before defining requirements
Use when verification has passed, to conduct project retrospective and capture reusable knowledge
Use when Delivery review has passed, to perform deep functional verification and acceptance testing
Use when needing to check project status, advance to next phase, handle escalations, or manage the project lifecycle state machine
Qiushi Skill: methodology skills for AI agents guided by seeking truth from facts, with Claude Code, Cursor, OpenClaw, Codex, OpenCode, and Hermes guidance.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Intelligent prompt optimization using skill-based architecture. Enriches vague prompts with research-based clarifying questions before Claude Code executes them