dy9759

Use when coordinating Claude Code + Codex as dual agents with Gemini as frontend/UI specialist consultant (via gemini-cli MCP), delegating parallel implementation of bounded modules to Codex, splitting work by boundary-clarity not domain, controlling token consumption, cross-harness setup (Cursor/Codex/OpenCode), pre-task thinking (/co-think), strategic plan review (/co-plan-review), engineering principles enforcement (Hyrum, Beyoncé, Chesterton, trunk-based, shift-left, feature-flags, deprecation, maintainability harness, testing quality), UI style standard (shadcn/radix-nova), and knowledge compounding. Self-contained; optional plugins and Gemini MCP auto-detected.

Pre-task thinking for claude-codex-orchestration (grill / office-hours style). Source-first: inspect repo/docs before asking. Product/Design, Technical/Approach, and Domain/Docs lanes. Ask only user-owned or genuinely ambiguous questions, one at a time, with recommendation + confidence. High-verification-risk fuzzy work becomes a Run Contract before dispatch. Always ends with Premise Challenge. Optional Codex cold-read second opinion. Delegates to references/thinking-decision.md.

Knowledge compounding for claude-codex-orchestration. Capture a solved non-trivial problem to docs/solutions/[category]/[slug]-[date].md with YAML frontmatter. Two modes: Full (4 parallel subagents — Context Analyzer / Solution Extractor / Related Docs Finder / Session Historian; overlap detection; Phase 2.5 selective refresh of older docs; discoverability check) or Lightweight (single pass). Delegates to claude-codex-orchestration references/knowledge-compounding.md.

Autonomous background refinement loop for claude-codex-orchestration. Runs /co-eval + /co-review, if candidate ≥ 6 runs /co-promote and commits, then schedules next cycle via ScheduleWakeup (270s, stays within cache TTL). Never self-stops — continues until user manually interrupts. Darwin ratchet: next session scoring the promoted dimension must improve or rule is reverted. Strategy escalation for persistent weak points (5/8/12/15+ sessions).

CEO-mode plan review for claude-codex-orchestration. Run after an Execution Plan is drafted. User chooses one of four modes (EXPAND/SELECTIVE/HOLD/REDUCE) dictating posture — dream bigger vs baseline+cherry-pick vs make-bulletproof vs minimum-viable. Seven Prime Directives apply in all modes: zero silent failures / named errors / shadow paths / interaction edge cases / observability-is-scope / deferred-is-written-down / security-is-scope. Critical rule: every scope change is explicit user opt-in. Delegates to references/thinking-decision.md.

Cross-session history search for claude-codex-orchestration. Before starting complex work, dispatch a general-purpose subagent to search ~/.claude/projects/ and ~/.codex/sessions/ for prior investigations of the same problem. Returns prior approaches tried, dead ends and why they failed, key decisions + rationale, related context. Avoids repeating failed paths from past sessions. Delegates to claude-codex-orchestration references/knowledge-compounding.md.

Layer 1 session self-eval for claude-codex-orchestration. Two-axis rating (Ambition × Execution) with locked 3×3 composite matrix (score 1–5), mandatory Devil's Advocate (argue lower AND higher before committing), anti-inflation check (last 5 entries, 4+ identical flags). Persists to .eval-scores.jsonl. Run at session end or when long-session prompt triggers (≥10 dispatches OR ≥6h since last eval). Delegates to claude-codex-orchestration references/self-correction.md Layer 1.

Layer 3 promote-to-SKILL for claude-codex-orchestration. Take a candidate identified by /co-review (Durability + Impact + Scope ≥ 6), distill from descriptive to prescriptive, write the rule into SKILL.md or appropriate reference file, remove source .error-log.jsonl entries, trigger Layer 0 auto-score and README sync. Ratchet rule enforces next same-dimension session must improve or revert. Delegates to claude-codex-orchestration references/self-correction.md Layer 3.

多平台编码助手。遵循各平台官方文档做编码规范、单测与编译/lint；协助将核心技术梳理为完整 WPS 笔记技术文档。生成的笔记必须包含 7 个二级标题（核心技术、核心代码、关键技术点、核心类和职责、调用链、架构概览、注意事项）；其中架构、核心技术、调用链的图示优先用 WPS 笔记的 generate_image 根据描述生成图片再用 insert_image 插入。用户新增标题时根据诉求补充内容；用户未关闭当前笔记期间约 1 分钟后主动更新直至关闭。当用户使用 Cursor、Codex、Claude Code、AS code 且提到架构、设计图、核心方法、关键技术或技术文档时，自动读写在 WPS 笔记。先 list_notes 先查后编；核心代码可从注释、复制、剪切板、选中或指定函数获取。子 skill review-notes 与 reference 负责流程细节。每30s监控一个笔记内容是否变动，如果变动自动更新文档。

将本地文档批量导入到 WPS 笔记。支持扫描 Obsidian Vault、思源笔记、微信公众号存档、 下载目录或任意用户指定目录，自动识别 HTML、Markdown、PDF、DOCX、PPTX、XLSX 等格式， 转换为 WPS 笔记可读内容并保留图片和富文本格式。 当用户说「导入文档到 WPS 笔记」「把我的 Obsidian 笔记导入」「导入思源笔记」 「把下载的 PDF/Word/PPT 导入笔记」「把公众号文章导入笔记」「同步本地文档到 WPS 笔记」时触发。 不适用于直接编辑 WPS 笔记内容、文档格式转换（不导入到笔记）。

基于 WPS 笔记知识库的新闻智能解读。将新闻存入笔记，搜索用户整个笔记库找到关联内容， 产出个性化 insight 分析。也支持批量新闻收集写入简报。 当用户说「找新闻」「热点汇总」「新闻简报」「帮我读这篇文章」「看看这个链接」 「这条新闻和我的项目有什么关系」或谈论/分享任何具体新闻资讯时使用。 不适用于：笔记的常规读写编辑、非新闻类内容的保存、纯知识问答。

笔记标签整理的核心原则与完整工作流程（CLI 版）。通过系统命令行调用 wpsnote-cli 操作 WPS 笔记，无需 MCP 服务连接。当用户提到"整理笔记标签"、"清理标签"、"标签太乱"、"标签太多"、"帮我打标签"、"重构标签"、"重新分类"、"笔记分类混乱"、"标签体系需要优化"等需求时使用。

笔记标签整理的核心原则与完整工作流程。当用户提到"整理笔记标签"、"清理标签"、"标签太乱"、"标签太多"、"帮我打标签"、"重构标签"、"重新分类"、"笔记分类混乱"、"标签体系需要优化"等需求时使用。

将网页内容高质量导入到 WPS 笔记，保留原文颜色、粗体、标题格式，图片按原文位置插入。 支持微信公众号文章、X/Twitter 推文/Thread 和任意通用网页，统一入口自动识别，加 --wps 参数直接写入 WPS 笔记。 当用户说「把这个网页存到笔记」「导入这篇文章」「抓取这个页面到笔记」 「把公众号文章存到 WPS 笔记」「把这条推文存到笔记」「收藏这个链接」「网页转笔记」时触发。 不适用于：新闻智能解读（用 news-to-note）、本地文档批量导入（用 doc-importer）。

【深度搜索】深挖笔记关联，构建知识图谱的 WPS 笔记查询助手。 当用户说"深度搜索""帮我深挖""关联查询""全面梳理"时使用。 支持跨笔记关联挖掘、语义发散扩展、知识沉淀，不同于简单关键词匹配。 不要用于单关键词查询（那用 search_notes MCP 工具）。

通过 MCP 工具读取、编辑和管理 WPS 笔记，基于 block 文档模型，所有内容以 XML 格式交换。当用户说"帮我看看笔记"、"搜索笔记"、"创建一篇笔记"、"编辑笔记内容"、 "整理标签"，或提到 WPS 笔记、WPS Note、云笔记时使用。也适用于排查 MCP 工具调用错误 （BLOCK_NOT_FOUND、EDITOR_NOT_READY 等）。

Comprehensive system architecture design and implementation workflow that orchestrates expert analysis, technical decision-making, and architectural pattern selection using the integrated toolset. Handles everything from initial system analysis to implementation-ready technical specifications.

Advanced code testing and review expert system that provides comprehensive code quality analysis, security vulnerability assessment, test strategy design, and quality assurance through multi-expert collaboration and intelligent tool integration.

Advanced context engineering management system that provides comprehensive context architecture design, memory management, knowledge engineering, and workflow orchestration through expert collaboration and intelligent tool integration.

Advanced frontend web development expert system that provides comprehensive modern web development services including architecture design, UI/UX implementation, performance optimization, engineering setup, and cross-platform development through expert collaboration and intelligent tool integration.

Complete PRD (Product Requirements Document) creation workflow that guides users from initial idea to implementation-ready specification using the integrated toolset in this repository. Handles everything from brainstorming and market analysis to technical requirements and team coordination.

Comprehensive LLM prompt engineering and optimization workflow that orchestrates expert analysis, advanced techniques, and multi-domain optimization using the integrated toolset. Handles everything from basic prompt improvement to complex multi-agent prompt orchestration.

Meta-skill creation and optimization system that analyzes, designs, and builds new specialized skills using the integrated toolset. Transforms skill creation from art to science through pattern recognition, intelligent tool selection, and automated quality assurance.

专业数据可视化专家，精通现代图表库、仪表板设计和交互式数据展示。能够将复杂数据转化为直观、美观且富有洞察力的可视化作品。

全面的DevOps和云原生专家，精通CI/CD流水线、容器化、Kubernetes部署和基础设施自动化。帮助企业实现现代化运维和持续交付。

React Native跨平台移动开发专家，精通iOS和Android应用开发、性能优化、原生模块集成和移动应用发布流程。能够构建高质量的移动应用。

自动化测试专家，精通单元测试、集成测试、E2E测试和性能测试。能够设计和实施全面的测试策略，确保软件质量和可靠性。

专业安全审计专家，精通漏洞扫描、渗透测试、合规检查和安全代码审查。帮助企业识别和修复安全风险，确保系统和数据的安全性。

Connect to AgentMesh network — first-time setup or daily dashboard with messages, agents, and natural language commands. Use when user wants to join, check, or interact with the agent mesh network.