基于费曼学习法的个人知识教练，通过"以教代学"的方式帮助用户深度理解和巩固知识。 支持主动知识回顾、薄弱点诊断、个性化学习建议，并可通过配置实现每日自动触发知识回顾。 触发命令：/feynman（直接调用费曼学习功能）
Tutors users from knowledge-base articles (Notion/local) using Feynman teach-back plus Sigma mastery gates, then writes Q&A talking points to a user-chosen location. Requires companion skills sigma and feyman-coach. Use when learning from KB/interview notes, Notion tutoring, Feynman+mastery prep, or /kb-feynman /费曼知识库 /面试口径学习.
Personalized 1-on-1 AI tutor using Bloom's 2-Sigma mastery learning. Guides users through any topic with Socratic questioning, adaptive pacing, and rich visual output (HTML dashboards, Excalidraw concept maps, generated images). Use when user wants to learn something, study a topic, understand a concept, requests tutoring, says 'teach me', 'I want to learn', 'explain X to me step by step', 'help me understand', or invokes /sigma. Triggers on: learn, study, teach, tutor, understand, master, explain step by step.
Designs Agent / LLM application architectures, multi-agent orchestration, tool-use pipelines, RAG / Text-to-SQL workflows, architecture route comparison, and architecture review. Extracts reusable architecture patterns from cookbooks, notebooks, and demo repos. Outputs solution-level plans (modules, data flow, agent patterns, MVP)—not product or tech-stack recommendations. Use when the user asks for 架构设计 / 方案设计 / Agent 工作流 / 多 Agent 编排 / 架构评审 / 方案路线 / MVP 路径, or says "design", "architecture", "review my design", "how should I structure this agent". Do NOT use for direct coding, bug fixing, product selection (framework/model/cloud), or pure API-question answering.
编排 Codex 的多 agent 自动化编码与交付流程，结合 AGENTS.md、.codex/config.toml、hooks、skills、MCP、CI 和 subagents，把复杂任务拆成 orchestrator、planner、explorer、docs_researcher、implementer、reviewer、release_guard 等角色协作。当用户需要设计或执行多 agent coding workflow、复杂功能开发、跨文件改动、PR 审查、发布前把关、把单 agent 任务升级为可控多 agent 流水线，或初始化当前仓库以接入这套工作流时使用。
检查 Skill 是否符合 Anthropic 官方最佳实践，执行自动化结构检查和 AI 内容质量审查，输出分级检查报告。当用户要求验证、检查、审查、评估某个 Skill 或 SKILL.md 是否规范时使用。
美化 nanobot 的回复输出格式，使用结构化排版、符号标记和清晰层级，适配 QQ、Telegram 等聊天平台。当需要让 nanobot 回复更美观、更易读时使用。
通过 gh CLI 搜索 GitHub 仓库、Issues、PRs，并管理本地知识库。用于查找 GitHub 资源、克隆仓库到知识库、维护项目目录。当用户搜索 GitHub、查找仓库、克隆项目、或维护本地代码知识库时使用。