Cursor 必备 MCP 组合推荐

以恰当的拆解方式引导重构、清理或重组。当用户要求重构、简化、抽取或现代化代码时使用。

Store and retrieve approved tenant-local team context with RuVector-backed isolation and provenance.

Remotion 最佳实践——在 React 中创作视频。包含 29 条领域专属规则，覆盖 3D、动画、音频、字幕、图表、转场等。当用 Remotion 在 React 中制作视频时使用，如涉及动画、音频、字幕、图表或转场。

掌握 ES6+ 特性，包括 async/await、解构、展开运算符、箭头函数、Promise、模块、迭代器、生成器与函数式编程模式，编写简洁高效的 JavaScript 代码。在重构遗留代码、实现现代模式或优化 JavaScript 应用时使用。

使用 Tailwind CSS v4、设计令牌、组件库与响应式模式构建可扩展的设计系统。在创建组件库、实现设计系统或标准化 UI 模式时使用。

使用 Express/Fastify 构建生产就绪的 Node.js 后端服务，落实中间件模式、错误处理、认证、数据库集成与 API 设计最佳实践。在创建 Node.js 服务器、REST API、GraphQL 后端或微服务架构时使用。

Design Video.js TypeScript APIs. Use for public shape, inference, state, extension points, middleware, composition, or ergonomics.

使用、配置、换肤、扩展与编排 Hermes Agent。

Maintain Supabase API types. Use when changing generated API type declarations, OpenAPI schemas, or investigating API type deployment drift.

Check secret handling, staged changes and push targets, destructive Git operations, Docker image hygiene, and new dependencies before consequential coding actions. Use for commits, pushes, container builds or publishes, dependency additions, and explicit safety reviews.

用于独立微应用：React Router SSR/SSG、Cloud overlay、Workers、CDN/R2、SEO/OG、网关路由、预览、Docker 集成与 SSR bundle 隔离。

Translate and synchronize application JSON locale files using source-key usage, project terminology, and focused validation. Use when adding keys or languages, updating source copy, or reviewing missing and stale translations.

Use when writing or changing a model-facing tool definition (name, description, parameter schema, or its system-prompt section), and when designing skills, context loading, or multi-step workflows, to make information discoverable and use context efficiently.

在 Kubernetes 环境中从 ingress-nginx 迁移到 Higress。适用于（1）分析现有 ingress-nginx 配置（2）读取 nginx Ingress 资源与 ConfigMap（3）通过 helm 安装 Higress 并正确设置 ingressClass（4）识别不受支持的 nginx annotation（5）为 nginx snippets/高级特性生成 WASM 插件（6）构建自定义插件并推送到镜像仓库。支持完整迁移工作流，含兼容性分析。

Guidelines on designing and building UI. Use when working in editor-ui or design-system packages. Triggers for tasks that include refactoring components, styling changes, or feature work.

Design or review Agent Builder interfaces. Use for Agent configuration patterns, modal flows, responsive behavior, status treatments, and other Agent-only UI work.

Authors and calibrates Instance AI evaluations that build standalone n8n Agents through Agent Builder. Use when a change under packages/cli/src/modules/agents affects build-agent routing, Agent setup, model or credential selection, tools, MCP servers, integrations, skills, tasks, testing, or user-facing build responses. Requires LangTracer access before authoring so each finished case can be published.

处理一次性操作：请求是一次具体且仅发生一次的效应——把数据导出或复制到某处、迁移、回填、清理——不含触发器、调度或复用意图。工作流只是载体而非交付物。用户很少会说一次性，需从任务形态推断。在为这类请求开始构建前加载；当构建工作流的结果中 postBuildFlow.reason 为 direct-one-off-build-succeeded 时也加载。用户会重复运行的自动化不要加载本技能——那走常规构建加 post-build-flow 路径。

通过两个判断对自动化请求分类：anchor（哪个原语掌握顶层控制流——工作流为锚、代理为锚、需澄清，或超出范围）与 embeds_other（另一原语是否内嵌出现——工作流中嵌入一个 agent 步骤，或代理把某个工作流当作工具调用）。每当当前轮次需要选定或重新判定某个自动化请求的意图时都必须使用，包括复合请求、构建过程中新引入的独立自动化、需要查询无法直接访问的外部系统的一次性提问或报告，以及在确定 anchor 前需要澄清的请求。当对话已明确面向某个工作流或 Agent 时，不要为常规编辑或扩展加载本技能。

使用 @PublicApiController 为 n8n Public API v1 新增、迁移或更新端点，涵盖公开 DTO、API key 与 RBAC scope、游标分页、OpenAPI 与覆盖率接线及测试。在 packages/cli/src/public-api/v1/ 下工作，或需要把已有服务通过 /api/v1 暴露时使用。

通过内部 Nathan 机器人从仓库（而非 Slack）部署临时的 n8n 测试实例，或生成一条本地 docker 运行命令。在开 PR 后向用户提供可实时测试的实例，或用户要求为某分支搭建/部署测试实例时使用。

在为新建或已有的 n8n Agent 调用 build-agent 前加载。负责前置资源创建、忠实承接用户请求、跨轮次的目标 agent 定位、向用户提问、测试与发布。当对话已明确面向某个 Agent 时，可直接用于常规跟进事项；只有当所需产物不再清晰时才重新执行意图识别。

指导在 packages/frontend/editor-ui 中开展实验的工作。在创建、扩展、接线、测试、评审或下线 editor-ui 实验，以及 PostHog 特性开关、实验 key 索引、变体、store/composable、实验持久化状态或实验埋点时使用。

Add or change Rust diagnostic tracing in litellm-rust, including route spans, subscriber layers, and Python logger delivery

使用 Plankton 在编写时强制执行代码质量——通过钩子在每次文件编辑时自动格式化、代码检查并由 Claude 驱动修复。适用于为文件编辑设置写入时格式化、检查或自动修复钩子。

Harbor intake for external work — the captain only handles unresolved decisions. Sweeps incoming issues and PRs, verifies every claim before disposition, reuses every decision already made, and hands the maintainer a docket whose pending items each carry one question with options, recommendation, impact and evidence. Agent-autonomous for facts and for actions covered by standing authorization; signed for every new judgment. Never merges.

在进行 ArchiveBox 应用、CLI、服务端、Docker 镜像、Admin UI、REST API、数据目录、爬取、快照以及发布与部署脚本相关工作时使用。

Answer questions about Everything Claude Code by reading the live repo surface — agents, skills, commands, hooks, rules, install profiles, and docs — instead of memory. Use when the user asks what ECC includes, how to install or reset it, which skill or command fits a task, or how project onboarding works.

Assemble scored competitor profile cards (from benchmark-methodology) into a decision-grade competitive report with landscape map, competitor tiers, benchmarking matrix, white-space analysis, strategic recommendations, and team alignment trigger questions. Use when presenting competitive findings to leadership or a board, writing a competitive landscape report, or as the final step of the competitive analysis pipeline.

Protected Health Information (PHI) and PII compliance patterns for healthcare applications: data classification, row-level access control, tamper-proof audit trails, schema tagging, and common leak vectors such as logs, URLs, and browser storage. Use when code touches patient or clinician data, when implementing HIPAA or GDPR access controls, or when auditing a healthcare system for data exposure.

Readiness checklist for homelab VLAN segmentation, local DNS filtering (Pi-hole, AdGuard Home), and WireGuard-style remote access. Use when planning or reviewing home network changes — splitting a flat network into trusted, IoT, guest, or management VLANs, moving DHCP to a local resolver, or adding VPN access — before changing router, firewall, DHCP, or VPN configuration.

Operate and extend NanoClaw, ECC's zero-dependency session-aware REPL, with persistent markdown-backed sessions and slash commands for model switching, skill loading, session branching, cross-session search, history compaction, and export. Use when running or extending scripts/claw.js, or when resuming, branching, compacting, searching, or exporting a NanoClaw session.

Run a structured, resumable multi-session brand identity interview across 8 modules (purpose, positioning, audience, personality, voice, narrative, founder tension) using laddering, 5 Whys, and projective techniques, persisting answers to disk and producing a master brandbook (90_SYNTHESIS.md). Use when creating or repositioning a brand, briefing designers or writers, or making implicit founder knowledge explicit.

Guard against destructive operations with three modes: Careful intercepts dangerous commands (rm -rf, git push --force, DROP TABLE) for confirmation, Freeze locks writes to one directory, and Guard combines both via PreToolUse hooks. Use when working on production systems, running agents autonomously, restricting edits to a directory, or during migrations, deploys, and data changes.

Run repeated rollouts ("Prime Gauss" style recursive prompting) while keeping an append-only decision ledger of trials, marks, coherence checks, and promotion gates, so recursive confidence never auto-approves live trading, deploy, or destructive actions. Use when the user asks for repeated rollouts, marked decision processes, high-dimensional search, stochastic optimization, local-optima exploration, ensemble comparison, or recursive reasoning with a visible evidence trail.

Measure performance baselines and detect regressions across browser Core Web Vitals (LCP, INP, CLS, page weight), API endpoint latency percentiles, and build/test feedback times, with before/after comparison stored in git-tracked .ecc/benchmarks JSON. Use when checking page speed, responding to 'it feels slow' reports, verifying launch performance targets, or comparing stack alternatives.

Validate the why before building through four product diagnostics — a YC-style product diagnostic that produces PRODUCT-BRIEF.md with a go/no-go recommendation, a founder review scoring product-market-fit signals, a user journey audit measuring time-to-value, and ICE feature prioritization. Use when pressure-testing product direction, choosing between features, sanity-checking a launch, or converting a vague idea into a product brief.

Ingest, index, search, edit, and monitor video and audio with the VideoDB Python SDK — upload from files, URLs, or RTSP feeds, build spoken and scene indexes with timestamped search and playable clips, transcode and reframe, do timeline edits (subtitles, overlays, dubbing), and run real-time alerts on live streams or desktop capture. Use when working with video search, transcription, clipping, transcoding, streaming, or live video alerts.

面向 Claude Code 会话的全面验证系统。在宣称任务完成前验证 Claude Code 会话的工作成果时使用。

面向 Claude Code 会话的正式评估框架，落实评估驱动开发（EDD）原则。当 Claude Code 工作流在获得信任或被修改前需要正式评估时使用。

将 Claude Code 会话中做出的架构决策捕获为结构化 ADR。自动检测决策时刻，记录上下文、考虑的替代方案与理由。维护 ADR 日志，帮助后来的开发者理解代码库为何如此构建。

把描述出来的工作流映射到合适的 ECC 命令组，给出执行顺序与停止条件，并可浏览全部命令组配方族。在扁平的命令目录之上增加了“按族分组 + 执行顺序 + 何时停止”这一层。仅提供建议。触发时机：用户问某件事该用哪些命令、哪个命令组负责它、查看 ECC 配方、列出 ECC 流水线，或问如何用 ECC 跑某个工作流。不触发：用户想直接执行任务、想要单条命令的详细文档（改用 ecc-guide），或想改写提示词草稿（改用 prompt-optimizer）。

在编写新功能、修复 bug 或重构代码时使用此 skill。强制推行测试驱动开发，覆盖率不低于 80%，涵盖单元、集成与 E2E 测试。

可用于生产的 Dart 与 Flutter 模式，覆盖空安全、不可变状态、异步组合、Widget 架构、主流状态管理框架（BLoC、Riverpod、Provider）、GoRouter 导航、Dio 网络、Freezed 代码生成与整洁架构。在编写或评审 Dart 与 Flutter 代码——状态、Widget、导航、网络或架构——时使用。

在完成复杂任务、出现失败，或需要复盘学到了什么之后使用。教你写出能提炼可复用模式的成长日志——而不是日记。

为 Claude Code 提供按项目持久化的记忆。会话启动时自动加载项目上下文，用 git 活动追踪会话，并写入原生记忆。命令运行确定性的 Node.js 脚本——行为在各模型版本之间保持一致。当项目需要让上下文在多次 Claude Code 会话之间保留、而不必每次重新解释时使用。

使用 firecrawl 与 exa MCP 进行多源深度研究。检索网页、综合发现，并交付带有来源归属的引用报告。在用户需要对任意主题做有证据、有引用的彻底研究时使用。

先检索再编码的工作流。在动手编写自定义代码之前，先搜索已有的工具、库和模式。会调用 researcher 代理。