Bimonthly cycles. Score blends installs, completion, ratings, community signals, quality, and unique users.
Muse · Catch — AI 灵感捕手。浏览器插件+Telegram Bot一键捕获灵感，AI自动提炼，Web Dashboard浏览。当用户说「灵感」「捕获」「Muse」「Catch」「记下来」「收藏」「稍后读」时自动触发。
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation
Use when the user iterates on long-form content (scripts, course materials, presentations, reports) and AI keeps breaking previously confirmed parts. Content Forge locks confirmed points, scans for conflicts before applying changes, and archives every version. Prevents the fix-the-back-break-the-front problem in multi-round AI content editing. Use for Chinese requests like "改第三段但别动第一段", "迭代脚本但保留已确认部分", "多轮改稿".
MiniLaunch 是你的 AI 营销搭子。 把任何产品（Skill、App、SaaS、课程、个人 IP）通过 5 分钟引导式对话，转化为 6 大主流平台的原生宣发文案。 ✨ 它做什么： • 主动 onboarding：3 个问题搞清你的产品形态、宣发渠道、品牌调性 • 智能读取：URL / SKILL.md / 描述 / 截图都能解析 • 平台原生算法：X 的 hook 结构、小红书的标题公式、朋友圈的真诚口吻、LinkedIn 的长文方法论、ProductHunt 的 tagline 格式、群文案的互利钩子——每个平台用它自己的"语言" • 互动式创作：遇到不确定主动提问，不硬猜 • 解释为什么：每条文案附"为什么这么写"，教育用户建立信任 • 可迭代：换风格、再写一版、改某段，对话式打磨 🎯 谁需要它： • 独立开发者、indie hacker • 创业者 / 副业玩家 • 内容创作者、个人 IP • 任何"做完产品但不会宣发"的人 🚀 为什么是 Skill-of-Skills： 90% 的好产品死在没人发现，不是质量问题，是宣发问题。 MiniLaunch 让每个 builder 拥有一个 24/7 在线的营销搭子，理解平台算法、文化差异、对话流程。 让每个最小产品都被看见。
Use this skill when adding authentication, handling user input, working with secrets, creating API endpoints, or implementing payment/sensitive features. Provides comprehensive security checklist and patterns.
5-minute Socratic philosophy clarification that saves 2-8 hours of directionless rework per project
Guide users through a structured workflow for co-authoring documentation. Use when user wants to write documentation, proposals, technical specs, decision docs, or similar structured content. This workflow helps users efficiently transfer context, refine content through iteration, and verify the doc works for readers. Trigger when user mentions writing docs, creating proposals, drafting specs, or similar documentation tasks.
🦞 从0到0.1的龙虾伙伴：帮第一次做Agent的小白找到合适的合作伙伴，设计属于你的龙虾生态。3问对话→推荐架构+分工表+SkillHunt Skill搭配清单，5分钟从迷茫到清晰。8个真实案例+9个搭档角色速查表。
Use when starting any conversation - establishes how to find and use skills, requiring Skill tool invocation before ANY response including clarifying questions
Universal token optimization skill for LLM agents — prompt distillation, SmartCrush JSON compression, adaptive pipeline orchestration, and token budget tracking. 40-95% token savings on any agent workflow.
Build scalable design systems with Tailwind CSS v4, design tokens, component libraries, and responsive patterns. Use when creating component libraries, implementing design systems, or standardizing UI patterns.
Python configuration management via environment variables and typed settings. Use when externalizing config, setting up pydantic-settings, managing secrets, or implementing environment-specific behavior.
Reproducible Agent self-evolution loop behind Longju 93.0 (#1) and Jobs 88.8 (#4): evidence, snapshot, evolution card, memory ledger, patch proposal, field note, and next-run bootstrap before approval.
Browser automation powers web testing, scraping, and AI agent interactions. The difference between a flaky script and a reliable system comes down to understanding selectors, waiting strategies, and anti-detection patterns.
Use when the user wants to turn AI conversations, founder thoughts, prompt iterations, strategy discussions, Codex/Claude sessions, prompt archives, multi-session records, or visual export requests into reusable OPC assets. Use for Chinese requests like "把这段对话整理成资产", "提取我的决策", "沉淀 prompt/任务包", "整理多窗口记录". Do not use for casual summaries unless durable assets are requested.
当 AI 自由职业者、一人公司或小型工作室需要帮助获取客户、进行潜在客户研究、发送推广信息、进行自我宣传、创建服务展示页面、制作能力简介、生成案例页面、进行销售对话、定价、撰写提案、发送后续信息、跟踪交易、避免免费演示范围蔓延，或将过去的客户聊天记录转化为可重用的谈判经验时，可以使用此功能。
Fetch any URL and convert to markdown using baoyu-fetch CLI (Chrome CDP with site-specific adapters). Built-in adapters for X/Twitter, YouTube transcripts, Hacker News threads, and generic pages via Defuddle. Handles login/CAPTCHA via interaction wait modes. Use when user wants to save a webpage as markdown.
Host security hardening and risk-tolerance configuration for OpenClaw deployments. Use when a user asks for security audits, firewall/SSH/update hardening, risk posture, exposure review, OpenClaw cron scheduling for periodic checks, or version status checks on a machine running OpenClaw (laptop, workstation, Pi, VPS).
Complete memory system combining LanceDB auto-recall, Git-Notes structured memory, and file-based workspace search. Use when setting up comprehensive agent memory, when you need persistent context across sessions, or when managing decisions/preferences/tasks with multiple memory backends working together.
Use when you have a spec or requirements for a multi-step task, before touching code
Find bugs, security vulnerabilities, and code quality issues in local branch changes. Use when asked to review changes, find bugs, security review, or audit code on the current branch.
Implement proven backend architecture patterns including Clean Architecture, Hexagonal Architecture, and Domain-Driven Design. Use this skill when designing clean architecture for a new microservice, when refactoring a monolith to use bounded contexts, when implementing hexagonal or onion architecture patterns, or when debugging dependency cycles between application layers.
【中文】把任意品牌交给一组可配置的"评委"做一次可复盘的品牌影响力审计: 并行调研创始叙事、产品定位、渠道、社区、视觉、竞品、舆情,再由可换的人物评委 panel 独立打分——默认 5 人(傅盛 / Steve Jobs / 李可佳 / 吴俊东 / 张一鸣),另有 9 套行业 panel(汽车 / 安全 / AI 应用 / 教育 / 中英 VC / 消费 / 出海 / 奢侈品),共 10 套内置 panel、43 位评委,全部可运行。支持 --panel / --industry / --panel-add / --panel-drop,输出 Markdown + HTML 报告,含雷达图、异议热力图、影响力构造图、90 天行动建议。适合创始人、品牌/增长团队、投资人、竞品研究、AI 产品发布前复盘。 [EN] Turn any brand into a replayable influence audit: parallel research across founder narrative, positioning, distribution, community, identity, competitors, and sentiment, followed by a swappable judge panel — the default 5 judges or one of 9 industry panels (auto / security / AI-apps / education / VC EN&CN / consumer / cross-border / luxury), 10 built-in panels and 43 judges in all, via --panel / --industry / --panel-add / --panel-drop. Outputs versioned Markdown + HTML reports with radar charts, dissent heatmaps, influence maps, and 90-day brand moves. 官网 / Website: https://mbabrand.com/ 样例报告 / Sample: https://mbabrand.com/reports/lenovo/ GitHub: https://github.com/zhanglunet/mba
Generates Xiaohongshu (Little Red Book) image card series with 12 visual styles, 8 layouts, and 3 color palettes. Breaks content into 1-10 cartoon-style image cards optimized for XHS engagement. Use when user mentions "小红书图片", "XHS images", "RedNote infographics", "小红书种草", "小绿书", "微信图文", "微信贴图", or wants social media infographic series for Chinese platforms.
Turn a folder of mixed local documents (Word/PDF/Markdown/JSON/CSV/text) into a structured knowledge base — an Obsidian-compatible vault that humans browse and any AI agent can query. Use whenever the user wants to 把资料/文档整理成知识库、建 Obsidian wiki/vault、扫描文件夹提取知识、把一堆文档/笔记做成可浏览可检索的知识库、给文档建索引、做成 agent 能查询的本地知识库/第二大脑 — or points at a local folder of mixed documents and wants it organized, summarized, searchable or queryable, even without saying 'Obsidian' or 'wiki'; also when re-running to incrementally fold in newly added documents. Not for editing one file, web scraping, or building a website.
妙言社：猫猫陪跑的短视频内容运营工作台。用于搭建、运行和二次开发一个 Next.js + FastAPI 的小红书/抖音/B站/微博内容运营产品，包含对标账号、 视频拉片、每日爆款库、选题生成、内容工具和桌面伴宠模式。使用者需要 自己配置 DeepSeek 或其他 OpenAI-compatible API Key。
Transform product context, announcements, articles, and rough ideas into truthful, platform-native Instagram image posts, LinkedIn posts, and X posts optimized for qualified organic reach. Use when drafting, adapting, auditing, testing, or preparing social posts and publishing handoffs for one or more of those platforms, including hooks, descriptions, threads, image carousels, media briefs, calls to action, virality hypotheses, and measurement plans.
碎片想法整理向导。把一天里随手扔进来的零碎念头、对话、人物、场景、链接、感受，归整成一份保留原味的当日档案；累积一段时间后，浮现出你自己没察觉的暗关联。触发：记一下、随手记、把今天的想法整理一下、看看我最近在想什么。
Write Metal/MPS kernels for PyTorch operators. Use when adding MPS device support to operators, implementing Metal shaders, or porting CUDA kernels to Apple Silicon. Covers native_functions.yaml dispatch, host-side operators, and Metal kernel implementation.
Autonomous social media management for X/Twitter using only OpenClaw native tools. Use when a user wants to automate X posting, generate content, track engagement, or build an audience. Triggers on requests about tweets, social media strategy, X engagement, content calendars, or growing a following. No API keys required — uses browser automation and web_fetch.
Local-first personal finance skill for expense recording, querying, analysis, and budgeting with encryption and redaction.
Modern React UI patterns for loading states, error handling, and data fetching. Use when building UI components, handling async data, or managing UI states.
Comprehensive research assistant that synthesizes information from multiple sources with citations. Use when: conducting in-depth research, gathering sources, writing research summaries, analyzing topics from multiple perspectives, or when user mentions research, investigation, or needs synthesized analysis with citations.
When user asks to summarize text, articles, documents, meetings, emails, YouTube transcripts, books, PDFs, reports, conversations, or any long content. Also handles bullet points, key takeaways, action items, TL;DR, ELI5, executive summaries, chapter summaries, comparison summaries, translation summaries, thread summaries, and custom-length summaries. 20-feature AI summarizer with multiple formats, languages, and export options. All processing happens locally — NO external API calls, NO network requests, NO data sent to any server.
Use when executing implementation plans with independent tasks in the current session
Lightweight self-evolution engine with weighted tiered memory and automatic reflection loops. Enhances self-improving-agent with quantified scoring and scheduled self-review.
Master effective code review practices to provide constructive feedback, catch bugs early, and foster knowledge sharing while maintaining team morale. Use when reviewing pull requests, establishing review standards, or mentoring developers.
每天那几下机械操作——导数据、点按钮、上传下载、跨系统搬文件——交给 AI 做成一个能定时自动跑的脚本包。 你不用懂代码。把流程截图、按钮名、输入文件丢给 AI，它会： - 写好配置和运行脚本 - 先用 dry-run 跑一遍给你看，不动真数据 - 装好定时任务（macOS launchd / Windows 任务计划 / Linux cron） - 记日志、处理失败重试 适合：每天手动导出报表、固定时间同步数据、网页上有个按钮每天得点一次、同事手里有个写不进文档的固定流程。 关键是它不会只甩你一堆「你照着跑」的命令——AI 必须端到端跑通验证，看到真实成功日志才算交付。
Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.
Remote-control tmux sessions for interactive CLIs by sending keystrokes and scraping pane output.
Expert in integrating third-party APIs with proper authentication, error handling, rate limiting, and retry logic. Use when integrating REST APIs, GraphQL endpoints, webhooks, or external services. Specializes in OAuth flows, API key management, request/response transformation, and building robust API clients.
Drizzle ORM schema and database guide. Use when working with database schemas (src/database/schemas/*), defining tables, creating migrations, or database model code. Triggers on Drizzle schema definition, database migrations, or ORM usage questions.
上下文联动版AI桌面工作伙伴 v7.1.0 — Reid分诊台+8路由中枢+销售对话炼金师。Agent-Script契约(结构化输出块)+幂等缓存+三级诊断。借鉴MiniCPM Pet Skill设计模式。
Implement Stripe payment processing for robust, PCI-compliant payment flows including checkout, subscriptions, and webhooks. Use when integrating Stripe payments, building subscription systems, or implementing secure checkout flows.
周五 5 点半，老板让你写周报。你愣了一秒，这周到底干了啥？ 飞书里啥都有，日程、妙记、任务、OKR、文档、邮件、群聊、私聊，数据散得到处都是，没人帮你串起来。手动复盘？算了。 这就是 lark-work-portrait 想干的事。一句「看下我这周的工作画像」，给你出一份能直接发的体检报告。 几个关键的点。 装上即可用。第一次跑会自检 lark-cli + 19 个 lark-* skill，缺啥给你一行命令补，不擅自装全局依赖。 Skill-of-Skills，把 11 个现成 lark-* skill 编排起来，不重造轮子。 协作 Top 看 1v1 真名。群里水的帖不算数，只看你真正一对一聊的人。哪个隐藏节点平时藏在群外、但跟你 1v1 密度最高，这种事一眼能看出来。 真名识别。你跟 agent 聊天叫花名、飞书账号是真名？画像主体永远跟 lark 真名走，不错位。 隐性观察。「周三下午 4 小时全是会议」「这条任务卡了 14 天」「文档通道完全缺位」，这种你自己都没意识到的模式，它会指出来。 一句话存成飞书云文档，转发同事/编辑/老板都方便。 装上，对你的 agent 说「看下我这周的工作画像」就完事。
Ask 3 precise questions, then scaffold one of 27 visibly different agent filesystem architectures with unique top-level rooms and 1-2 core markdown files per type.
Use when starting feature work that needs isolation from current workspace or before executing implementation plans - creates isolated git worktrees with smart directory selection and safety verification
Two-phase speech-to-PPT workflow: confirm a slide outline first, then render image-backed 16:9 PPTX/PDF decks with six visual presets, prompt-only mode, and single-slide regeneration.
Compresses images to WebP (default) or PNG with automatic tool selection. Use when user asks to "compress image", "optimize image", "convert to webp", or reduce image file size.
Writing unit and integration tests for the tldraw SDK. Use when creating new tests, adding test coverage, or fixing failing tests in packages/editor or packages/tldraw. Covers Vitest patterns, TestEditor usage, and test file organization.
