Recently Added
Fresh skills just added to the collection. Be the first to try them out!
figma-live-compare skill
by ahamedmansoor1988
The Loupe endpoint is at https://getloupe.vercel.app/api/extract-styles.
espresso
by mzuleta6
Claude without the noise. Eliminates hedging, filler, and re-contextualization from every response — without ever sacrificing precision.
ixl-quiz-generator
by kdneedsaclaude-777
Generates structured, adaptive math quizzes for Grades 4–8 aligned exactly to the IXL Ontario skill taxonomy with precise skill codes (e.g. C.3, D.7). Use this skill whenever anyone asks to generate a quiz, create practice questions, build a question set, or produce curriculum-aligned problems for Grades 4–8 Math. Also triggers for: IXL skill practice, adaptive quizzes, student weak area focus, parent-configured learning paths, tutor-reviewed grade progression, or any request like "make Grade 6 fractions questions" or "quiz on G7 integers." Always use this skill — never generate quiz questions without it.
zora-cli
by sigma99000
The agent's full interface to Zora — the onchain social platform on Base — through the Zora CLI (npx @zoralabs/cli), for both first-time setup AND everyday use. Use it to stand up an identity (Zora profile, Coinbase Smart Wallet, creator coin, first post), and just as much to act on Zora afterward: buy and sell creator coins and post coins, browse what's trending, look up a coin's price, holders, or trades, check balances and holdings, send ETH or tokens, and read and reply to DMs. Trigger on anything Zora-on-Base — setup phrasings like "set me up on Zora", "make me a Zora account", "become an agent on Zora", but equally everyday ones like "buy this coin on Zora", "what's trending on Zora", "check my Zora balance", "look up on Zora", "sell half my Zora position", or "reply to my Zora DMs" — even when the user never names the CLI.
coding-quality-loop
by zaingz
"Use when a coding agent must turn a software goal, bug, issue, or refactor into a small, verified, independently reviewed code change."
lqe-translator
by tzwkb
LQE scoring + self-iteration agent for AI/human game-localization translations — ZH source to any target language (EN/TH live; project profiles nrc/wwm). Deterministic pre-check (23 builtin rules + project custom) finds mechanical errors; Claude judges semantic errors AND provides corrected translations in one pass; Python calculates score. /loop drives iteration until score ≥ 98. Triggers: LQE/LQA, 译文质检/评估/打分, translation QA.
tiny.place
by tinyhumansai
The social economy for autonomous AI agents.
scenery
by scenery-sh
Use when building, running, debugging, inspecting, validating, or generating clients for scenery applications. scenery is a Go-native service runtime and CLI using .scenery.json/.config.json app config, //scenery directives, typed endpoints, local dev supervision, logs, traces, metrics, workers, and TypeScript client generation.
llm-architecture-tracker
by tohnee
Continuously research, analyze, and report on the world's leading large language models — their architecture, capabilities, and pricing. Use this skill whenever the user wants to track, compare, survey, or deep-dive into frontier LLMs (OpenAI/GPT, Anthropic/Claude, Google/Gemini, DeepSeek, GLM/Z.ai, Kimi/Moonshot, Xiaomi MiMo, Qwen, StepFun, Mistral, Llama, NVIDIA Nemotron, MiniMax, and others). Trigger it for requests like "update me on the latest models", "compare model X and Y architectures", "generate an LLM landscape report", "what changed in DeepSeek V4", "build me a model comparison", "track new releases", "what are the top models on OpenRouter", or any request that involves collecting model technical reports / HuggingFace cards / ModelScope / OpenRouter usage + pricing and turning them into a rigorous architecture-and- capability analysis. Also use it to produce the recurring HTML research report. Use this even when the user only names one model or asks a seemingly simple "what's new" question — the value is in the systematic, sourced, architecture-level analysis this skill enforces.
working-with-agents
by pedrosekine
Field-tested practices for getting reliable results out of agentic AI coding tools (Claude Code, Cursor, Codex, and similar). Use this whenever you are about to delegate a non-trivial coding or building task to an AI agent, when an agent's output is breaking things or drifting from what you asked, when you are setting up rules/guardrails for a project, or when you want to decide how much of a task to hand off versus keep. Consult it even if the user just says "the agent keeps messing this up" or "how should I prompt this" — the failure modes it addresses are usually upstream of the prompt itself.
notebooklm
by kingqiu
Use when the user mentions NotebookLM, shares a NotebookLM URL, asks to query uploaded documents, wants source-grounded answers from a notebook, or wants to manage a local NotebookLM library from Codex.
stop-slop
by iamneongo
Remove AI writing patterns from prose. Use when drafting, editing, or reviewing text to eliminate predictable AI tells.
agent-team
by AidenXu-1
Use when the user wants multi-agent or multi-session collaboration, 多Agent协作, 多会话开发, 部门分工, 三层架构, 管理层/执行层/审核层, an "agent company", or needs to diagnose which departments a project should have.
ip-diagram-creator
by haloshin
个人 IP 图解创作器。用你的 IP 形象，生成专属内容图解。用于把用户授权提供的照片、主页截图、简介、账号资料或角色档案转成稳定的极简手绘图解角色，并继续生成长文配图、图文号知识卡、手机海报、方法拆解图、分镜故事图、IP 演讲 PPT、直播分享课件和全图化手绘 PPT 页面等内容视觉。它强调 IP 主角 + 执行 Agent 的知识图解系统：用户 IP 做主讲、拆解、批注和调度，执行 Agent 做搬卡片、反馈、试错、标记风险和递交结果。用户想做个人 IP 形象、用自己的形象做知识图、图文号知识卡、文章配图、课程图解、内容海报、方法拆解图、IP 演讲 PPT、直播分享课件，或上传照片/主页截图/简介希望做内容视觉时，使用本 Skill。
animusflow
by samsoft111
Build, architect, and extend AnimusFlow — an AI-native CMS SaaS platform for creating intelligent websites globally. Use this skill whenever the user mentions AnimusFlow, AnimusFlowStudio, asks to build CMS features, create themes, develop plugins, design the admin panel, work on multi-tenancy, integrate AI modules (content generation, chatbot, SEO, personalisation, analytics), implement billing/plans, or build the marketplace. Also trigger for database schema, API design, Laravel backend, Vue 3 + Inertia admin SPA, Blade site rendering, or any work scoped to either the AnimusFlow CMS or the AnimusFlowStudio creator platform.
browser-cli
by lexmount
Operate Lexmount remote browser sessions through the browser-cli command line tool. Use when Codex or another agent needs to create, list, inspect, keep alive, or close Lexmount browser sessions; manage persistent browser contexts, pick reusable contexts, or detect locked contexts; guide authentication with auth status/export-env/login; verify installation, environment, and API connectivity with doctor; open pages, wait for selectors, URLs, load state, network idle, text, or form values, click, click indexed selector matches, type, focus, blur, clear, inspect form fields, inspect element state and geometry, set form values, check or uncheck labeled controls, dispatch common DOM events, submit forms, navigate history, read or mutate localStorage/sessionStorage and document.cookie-visible cookies, screenshot, evaluate JavaScript, inspect interactive elements, or snapshot page title, URL, HTML, and body text through the CLI; or verify Lexmount browser credentials without writing custom Playwright code.
token-cost
by EvilFreelancer
Use when the user invokes /token-cost or asks for the floor / bedrock / self-cost / "real" price of LLM tokens on a self-hosted server — the physical lower bound from electricity plus hardware amortization, NOT the market API price. The skill measures power draw and generation speed ON the server (nvidia-smi, streaming benchmark) and asks for electricity tariff, hardware cost and amortization period, then prints cost per 1M input/output tokens. Triggers: "сколько на самом деле стоит токен", "цена самохоста", capacity and break-even planning.
auto-empirical-research-skills
by brycewang-stanford
Route empirical-research requests through the Auto-Empirical Research Skills catalog when this whole repository is installed as one skill in Codex, CodeBuddy, Claude Code, or another IDE. Use to choose and load the right vendored AERS skill for causal inference, econometrics, replication, manuscript writing, citation checking, de-AIGC editing, or full empirical-paper workflows without reading the entire repository at once.
blueprint
by tanishraj
A brief description of what this skill does
prototype-loop-orchestrator
by hw00913-bot
总控 PM 到演示原型的自动化工作流。Use when Codex/claude code needs to orchestrate a PM-to-demo prototype loop, including requirement normalization, project decomposition, create-prototype-project initialization, iterative implementation, verification, and manual annotation prompt preparation. 适用于原型自动化、项目拆分、演示 Demo 生成、手动标注提示词准备和静态前端原型交付。
devforge
by JefiSun
End-to-end web development pipeline. Use this skill whenever: a .docx BRD is provided with a build request, user says "build from BRD / implement this spec / run the pipeline / develop this feature / re-run feat-X", or a multi-phase web development workflow is needed. Stack-agnostic — supports nextjs14, react-vite, and more via stack files. Works for both new and existing repos.
hiranya-spy-edge-tracker
by hiranyaw
Complete reference for Hiranya SPY Edge Tracker v1.0 — trendline break detection system for SPY 1M 0DTE scalping. Use this skill for anything related to: the Pine Script indicator, trendline break detection, ATR stop loss, TP1/TP2 targets, corner table, bar outcome labels, exit conditions ($ADD / 9 EMA / QQQ), webhook setup, Supabase edge_log table, Flask webhook routes, or the spy-dashboard integration. Triggers on: "trendline break", "edge tracker", "Pine Script", "SL/TP", "exit signal", "webhook", "edge_log", "corner table", "bar label", or any question about building or improving this indicator system.
tossinvest-web-api
by dd3ok
Use this skill when the user asks for public, read-only TossInvest/토스증권 market data visible on tossinvest.com, including Korean/US stock quotes, order books, candles, financials, filings, news, rankings, screeners, calendars, indices, FX, exchange-rate widgets, crypto-like index pages, or public endpoint re-verification. Do not use for login, accounts, holdings, orders, authenticated broker workflows, bulk scraping, or investment advice.
llm-architecture-tracker
by tohnee
Continuously research, analyze, and report on the world's leading large language models — their architecture, capabilities, and pricing. Use this skill whenever the user wants to track, compare, survey, or deep-dive into frontier LLMs (OpenAI/GPT, Anthropic/Claude, Google/Gemini, DeepSeek, GLM/Z.ai, Kimi/Moonshot, Xiaomi MiMo, Qwen, StepFun, Mistral, Llama, NVIDIA Nemotron, MiniMax, and others). Trigger it for requests like "update me on the latest models", "compare model X and Y architectures", "generate an LLM landscape report", "what changed in DeepSeek V4", "build me a model comparison", "track new releases", "what are the top models on OpenRouter", or any request that involves collecting model technical reports / HuggingFace cards / ModelScope / OpenRouter usage + pricing and turning them into a rigorous architecture-and- capability analysis. Also use it to produce the recurring HTML research report. Use this even when the user only names one model or asks a seemingly simple "what's new" question — the value is in the systematic, sourced, architecture-level analysis this skill enforces.