Explore AI Agent Skills & Claude Prompts
Discover open-source agent skills for Claude Code, Codex, ChatGPT, and any tool that uses SKILL.md.
Enter through keywords, occupations, creators, and GitHub sources to see what kinds of skills are emerging across domains.
Use the same catalog through the API
Connect 381,784 public skills to your own search, analytics, or agent workflow with the REST API.
Querying local SQLite index...
payuni-upp-v2
by zenbuappsPAYUNi UPP V2 (UNiPaypage Version 2.0) 統一金流整合支付頁完整技術參考。 涵蓋 AES-256-GCM 加解密、SHA256 HashInfo、UPP 請求/回傳參數、所有付款方式 (信用卡/分期/紅利/Apple Pay/Google Pay/Samsung Pay/銀聯/ATM 虛擬帳號/超商代碼/ icash Pay/AFTEE/LINE Pay/街口支付/貨到付款/宅配)、信用卡 Token / 買方 Token、 優惠券、物流參數、交易查詢/請退款/取消授權、Token 查詢/取消、Sandbox 測試。 當程式碼涉及 PAYUNi、統一金流、payuni、UPP、UNiPaypage、台灣金流、 EncryptInfo HashInfo、信用卡分期、超商代碼、ATM 虛擬帳號、LINE Pay、 街口支付、AFTEE、icash Pay、Apple Pay、Google Pay、Samsung Pay、 /api/upp、/api/trade/query、/api/trade/close、/api/trade/cancel、 /api/credit_bind、/api/cancel_cvs、/api/aftee_direct 等端點時必須使用。 V2 專用 (Version 2.0、AES-256-GCM)，PAYUNi 官方目前 UPP 僅有 V2，沒有 V3。 與舊版 V1.x (AES-256-CBC) 不相容。
ezpay-invoice
by zenbuappsezPay 電子發票（藍新金流 NewebPay 旗下品牌）API 完整技術參考，對應官方文件 EZP_INVI_1.2.1「電子發票技術串接手冊（標準版）」（簡單行動支付股份有限公司， 2021-04-01）。台灣電子發票服務商，涵蓋發票全生命週期：開立發票 （即時 / 等待觸發 / 預約自動，B2C / B2B，invoice_issue）、觸發開立發票 （invoice_touch_issue）、作廢發票（invoice_invalid）、開立折讓 （allowance_issue）、觸發確認 / 取消折讓（allowance_touch_issue）、 作廢折讓（allowanceInvalid）、查詢發票（invoice_search）。包含 AES-256-CBC 加密 PostData_（HashKey 32 bytes / HashIV 16 bytes / PKCS#7 padding / hex 輸出）、回應 CheckCode SHA256 驗證機制、載具類別 （手機條碼 0 / 自然人憑證 1 / ezPay 電子發票載具 2）、捐贈碼 LoveCode、 課稅別 TaxType、零稅率報關標記、含稅 / 未稅金額計算邏輯、完整錯誤碼表、 測試與正式環境 endpoint。Use this skill whenever code or tasks involve ezPay、ezpay、藍新電子發票、電子發票、cinv.ezpay.com.tw、inv.ezpay.com.tw、 開立發票、發票作廢、折讓單、開立折讓、作廢折讓、查詢發票、PostData_、 MerchantID_、CheckCode、HashKey、HashIV、CarrierType、載具、LoveCode、 捐贈碼、TaxType、課稅別、invoice_issue、invoice_touch_issue、 invoice_invalid、allowance_issue、allowance_touch_issue、allowanceInvalid、 invoice_search、e-invoice integration、台灣電子發票串接，或在 zenbu-site apps/api-gateway/src/commerce/ 下整合 ezPay 電子發票開立。 本 SKILL 為唯一官方 API reference
newebpay-mpg
by zenbuappsNewebPay MPG (Multiple Payment Gateway) complete payment gateway reference for Taiwan, aligned with official NDNF-1.2.2 manual (revised 2026-04-21). Covers MPG Version 2.0 (legacy) and 2.3 (current spec): AES-256-CBC + SHA256 encryption, optional AES-256-GCM (EncryptType=1), all TradeInfo parameters, payment methods (credit card / WebATM / VACC / CVS / BARCODE / CVSCOM / LINE Pay / Apple Pay / TWQR / smart-ATM 2.0 / AFTEE BNPL / e-wallets), callback handling, query trade, credit card close/cancel/refund, e-wallet refund, BNPL refund (NPA-B07) and BNPL settle (NPA-B62), error codes, sandbox endpoints. Use this skill whenever code involves newebpay, TradeInfo, TradeSha, EncryptData_, HashData_, MerchantID + HashKey + HashIV, ccore.newebpay.com / core.newebpay.com (and legacy spgateway.com aliases), MPG/mpg_gateway, /API/QueryTradeInfo, /API/CreditCard/Close, /API/CreditCard/Cancel, /API/EWallet/refund, /API/Bnpl/refund, /API/Bnpl/settle, NewebpayService, newebpay callback/notify handling, AFTEE 先買後付, TWQR 跨機構支付
payuni-uni-embed-v3
by zenbuappsPAYUNi UNi Embed (免跳轉支付元件 / 內嵌式金流) API Ver 3.0 + JS SDK Ver 2.0 完整技術參考。 iframe 嵌入式信用卡收單，站內付不跳轉，後端 token_get 取 SDK_TOKEN，前端 uniPayment.js 收集卡片資訊取得綁定結果，後端再呼叫 merchant_trade 完成幕後授權。 當程式碼涉及 PAYUNi、統一金流、payuni-uni、UNi Embed、內嵌金流、站內付、tokenization、 iframe 信用卡收單、SDK 整合、uni-payment.js、createSession、getTradeResult、 iframe/token_get、iframe/merchant_trade、IFrameDomain、SDK_TOKEN 時使用。 與 payuni-upp-v3（導轉式整合支付頁）平行；加密規則（AES-256-GCM + SHA256 HashInfo） 與 UPP 完全一致，請參考 payuni-upp-v3 該段。本 SKILL 為 V3 (2025/09 釋出)， 若程式碼出現 V2 或 V1 路徑，請勿混用——V3 token_get 階段不送訂單資料。
antd-toolkit
by zenbuappsantd-toolkit (j7-dev/antd-toolkit) v1.3.x — React component library wrapping antd v5 with Refine integration helpers, WordPress / WooCommerce specific UI, BlockNote-based rich-text editor, and a Bunny Stream media library. Use this skill whenever code imports from 'antd-toolkit', 'antd-toolkit/wp', or 'antd-toolkit/refine', or when package.json contains "antd-toolkit". Triggers include LocaleProvider, EnvProvider, ActionButton, ObjectTable, SimpleModal, SimpleDrawer, BlockNote, MediaLibrary, BunnyProvider, dataProvider, bunnyStreamDataProvider, useWoocommerce, ProductFilter, UserAvatarUpload, notificationProvider, useColumnSearch, and the WP / WooCommerce constant sets (POST_STATUS, ORDER_STATUS, PRODUCT_TYPES). Do NOT search the web for antd-toolkit docs — use this skill instead.
line-login-sdks
by zenbuappsLINE Login SDKs official reference at API-reference depth. Covers the native mobile/game SDKs that integrate LINE Login (LINE's OAuth 2.0 / OpenID Connect social login) into apps: the LINE SDK v5 for iOS Swift, the LINE SDK v5 for Android, the LINE SDK for Unity, and the LINE SDK for Flutter plugin. Use this skill whenever the task touches client-side LINE Login on a native app: installing the SDK (CocoaPods `LineSDKSwift`, Carthage, Maven Central `com.linecorp.linesdk:linesdk`, Unity `.unitypackage`, pub.dev `flutter_line_sdk`), configuring channel ID / bundle ID / package name / URL schemes, calling LINE Login, getting an access token / ID token, getting the user profile, logging users out, refreshing or verifying access tokens, enabling the add friend option (bot prompt) and checking friendship status, handling SDK errors, using universal links on iOS, or using the SDK from Objective-C. Also includes the full per-platform SDK class/method API reference — every class, interface, enum, property, method signa
cloudflare-pages-wrangler
by zenbuappsCloudflare Pages + Wrangler CLI 完整技術參考（v3+/v4）。 當任務涉及以下情況時，必須使用此 skill： 使用 wrangler CLI（pages deploy/dev、deploy、dev、tail、secret put/delete、types、versions、rollback）； 撰寫或修改 wrangler.toml、wrangler.jsonc、wrangler.json； GitHub Actions 中 uses: cloudflare/wrangler-action@v3、uses: cloudflare/pages-action； Cloudflare Pages Functions（functions/ 目錄、onRequest、onRequestGet/Post/Put/Delete/Patch/Options、 _middleware.ts、_worker.js、[param]、[[catchall]]、context.env、context.params、context.request、context.next、context.waitUntil）； _routes.json、_headers、_redirects、pages_build_output_dir； Pages bindings（KV/R2/D1/Durable Objects/Queues/Services/Hyperdrive/Vectorize/AI/Browser）； 部署到 Cloudflare Workers（main、compatibility_date、compatibility_flags、routes）； Pages 建置設定（build command、build output directory、CF_PAGES_* 系統環境變數）； 從 Cloudflare Pages 遷移到 Cloudflare Workers（Workers Static Assets）。 工具：wrangler ^3.x/^4.x、cloudflare/wrangler-action@v3。
refine-v5
by zenbuappsRefine framework v5 的完整技術參考。Refine 是 headless 的 React meta-framework， 用於快速建構 CRUD 應用（admin panel、dashboard、B2B 工具、內部後台）。 涵蓋 @refinedev/core 所有 hooks、components、5 大 provider（data / auth / router / notification / live / access-control / audit / i18n）、Ant Design 與 shadcn/ui 整合， 以及 v4 → v5 完整遷移注意事項。 當程式碼涉及 Refine 或相關 import 時，必須使用此 skill 而不是去搜尋 web。 即使用戶沒有明確說出「Refine」，只要任務涉及以下關鍵字也應使用此 skill： @refinedev/core、@refinedev/antd、@refinedev/react-hook-form、@refinedev/react-table、 @refinedev/simple-rest、@refinedev/react-router、<Refine>、useTable、useForm、useList、 useOne、useCreate、useUpdate、useDelete、useShow、useSelect、useMany、useInfiniteList、 dataProvider、authProvider、routerProvider、liveProvider、useGo、useResourceParams、 useMenu、useLogin、useLogout、useCan、refineCore、tableProps、formProps、saveButtonProps。 此 skill 提供的資訊對應 Refine v5（@refinedev/core ^5.x、@refinedev/antd ^6.x）， v5 移除了 v4 的所有 deprecated API、升級至 TanStack Query v5、支援 React 18/19， 與 v4 有大量 breaking changes（return type 重構、metaData→meta、queryResult→query 等）， v4 的 snippet
antd-v5
by zenbuappsAnt Design v5 (antd ^5.x) complete API reference for React component library. Use this skill whenever the task involves any antd v5 component: Table, Form, Select, Input, Button, Modal, notification, message, Upload, Tag, Tooltip, Badge, Image, Radio, Segmented, Switch, DatePicker, RangePicker, Popconfirm, Drawer, Descriptions, ConfigProvider, Space, Layout, Grid (Row/Col), Spin, Pagination, Menu, Dropdown, Breadcrumb, Tabs, Steps, Affix, Anchor, AutoComplete, Cascader, Checkbox, ColorPicker, InputNumber, Mentions, Rate, Slider, TimePicker, Transfer, TreeSelect, Avatar, Calendar, Card, Carousel, Collapse, Empty, List, Popover, QRCode, Statistic, Timeline, Tree, Alert, Progress, Result, Skeleton, Watermark, Divider, Flex, FloatButton, Splitter, Typography, App, or the antd theme system (token, algorithm, useToken, CSS variables). Also use when code imports from 'antd', references antd TypeScript types (TableProps, FormInstance, ColumnsType, UploadFile, etc.), or when configuring antd theming/tokens. Do NOT use
antd-v6
by zenbuappsAnt Design v6 (antd ^6.x) complete API reference for React component library. Use this skill whenever the task involves any antd v6 component: Table, Form, Select, Input, InputNumber, Button, Modal, Drawer, notification, message, Upload, Tag, Tooltip, Badge, Image, Radio, Segmented, Switch, DatePicker, RangePicker, Popconfirm, Descriptions, Tabs, Menu, Steps, Timeline, ConfigProvider, Space, Space.Compact, Layout, Grid (Row/Col), Masonry, BorderBeam, or the antd v6 theme system (token, algorithm, useToken, cssVar, zeroRuntime). Also use when code imports from 'antd', references antd TypeScript types (TableProps, FormInstance, ColumnsType, SelectProps, UploadFile, etc.), uses the v6 semantic DOM API (classNames / styles props with semantic keys like root / popup.root / body), or migrates an antd v5 project to v6. Also covers the Ant Design tooling: the @ant-design/cli command-line tool (antd list / info / doc / demo / token / doctor / lint / migrate) and the official Ant Design MCP server (antd mcp) for IDE /
wp-phpstan
by zenbuappsUse when configuring, running, or fixing PHPStan static analysis in WordPress projects (plugins/themes/sites): phpstan.neon setup, baselines, WordPress-specific typing, and handling third-party plugin classes.
react-flow-v12
by zenbuappsComplete API reference for @xyflow/react (React Flow) v12+. Use this skill whenever any task involves @xyflow/react, ReactFlow, node-based UI, flow diagrams, DAG visualization, graph editors, workflow builders, canvas editors, or any code importing from '@xyflow/react'. Covers ReactFlow component props, useReactFlow hook, useNodes, useEdges, useNodesState, useEdgesState, useConnection, useViewport, useNodeId, useHandleConnections, useNodeConnections, useNodesData, useNodesInitialized, useOnSelectionChange, useOnViewportChange, useUpdateNodeInternals, useKeyPress, useStore, useStoreApi, useInternalNode hooks; Background, Controls, MiniMap, Panel, Handle, NodeToolbar, NodeResizer, EdgeLabelRenderer, EdgeToolbar, BaseEdge, EdgeText, ViewportPortal components; addEdge, applyNodeChanges, applyEdgeChanges, getBezierPath, getSmoothStepPath, getStraightPath, getSimpleBezierPath, getConnectedEdges, getIncomers, getOutgoers, getNodesBounds, getViewportForBounds, reconnectEdge, isNode, isEdge utils; Node, Edge, NodeProp
Browse Agent Skills by Occupation
23 major groups · 867 SOC occupations
Browse by Category
Explore agent skills organized by their primary use case
Explore the agent skills ecosystem by occupation and creator
SkillMD is not just a keyword search box. It is an open map that organizes public skills by occupation, creator, and repository, helping you see which workflows, judgment criteria, and domain habits people are writing for AI agents.
Then follow creators and GitHub repositories back to the source: compare the skills a team maintains, whether the repo is active, and how the README frames the work before you open, install, or reuse anything.
Use it three ways: learn an unfamiliar field by occupation, study how creators organize skills, then use source context to decide what is worth opening or reusing.
01 Map a field
Browse 23 occupation groups and 867 SOC roles to learn what skills exist in adjacent domains and how they break down real work.
02 Follow creators
Use creator and repository pages to inspect maintained skill collections, recent updates, and source context before trusting a result.
03 Search with sources
Search 1.7M+ collected skills, then use occupation tags, creators, and GitHub source context to decide what is worth opening.
Start with the occupation map, then follow creators and repositories back to real code. SkillMD helps explain why a skill is worth opening, not only what it is named.
Standardizing Agent Capabilities with SKILL.md and Model Context Protocol (MCP)
In the rapidly evolving landscape of artificial intelligence, LLM agents (Large Language Model agents) have transitioned from simple text predictors to autonomous problem solvers. To orchestrate complex, multi-step agentic workflows, developers require a standardized format to specify agent capabilities, prompt instructions, system rules, and database bindings. This is where SKILL.md and the Model Context Protocol (MCP) have emerged as standard developer paradigms. SkillMD serves as the central directory for indexing, exploring, and sharing these critical agent configurations.
Our open-source registry currently tracks over 1.7 million collected SKILL.md configurations and system prompts. By compiling agent configurations from active developers on GitHub, we bridge the gap between prompt engineering research and production execution. Whether you are building agents with Anthropic's Claude Code, OpenAI's GPT-4, Google's Gemini, or local models using Ollama and LlamaIndex, standardized skill definitions ensure your agents behave predictably across different runtime environments.
What is the Model Context Protocol (MCP)?
The Model Context Protocol (MCP) is an open-source standard designed to connect LLMs to data sources, developer tools, and external environments. MCP establishes a bidirectional communication channel between client applications (like Cursor, Claude Desktop, or custom agent systems) and servers hosting data or capabilities. Standardizing instructions via SKILL.md enables LLMs to query databases, read local files, execute terminal commands, and integrate third-party APIs. SkillMD allows you to find ready-to-run MCP servers and prompt instructions for various occupations and technical tasks.
The Structure of a Professional SKILL.md File
A valid SKILL.md configuration is designed to be easily read by humans and parsed by LLMs. It contains precise system instructions, trigger conditions, required parameters, and execution examples. Below is the typical architectural blueprint of a professional agent skill:
- Metadata & Core Scope: Declares the name of the skill, author details, target models, and a description of the capability.
- Triggers & Intent Detection: Details semantic triggers that help the agent decide when to invoke this skill.
- System Prompts: Explicit system-level instructions that direct the agent's behavior, personality, safety guardrails, and formatting preferences.
- Capabilities & Tools: Lists the files, databases, or APIs the agent must access to complete the tasks.
- Few-Shot Examples: Demonstrates real inputs and outputs, helping the model generalize behavior through in-context learning.
Optimizing Agent Workflows for Modern LLMs
Writing effective agent skills requires deep knowledge of prompt engineering. With the release of advanced reasoning models like Claude 3.5 Sonnet, ChatGPT o1, and DeepSeek-V3, prompt templates must focus on structured thinking. Developers are encouraged to use XML tags (e.g., <thought>, <context>, and <rules>) to isolate execution boundaries. Standardized prompts prevent agents from suffering from context drift, ensuring that long-running tasks remain aligned with the initial system parameters.
Exploring by SOC Occupations and Creator Profiles
What makes SkillMD unique is its taxonomy. Instead of simple text search, we parse and organize files according to the Standard Occupational Classification (SOC) system. This means you can discover skills written for Computer and Mathematical roles, Business and Financial operations, Legal, Design, and and Educational Instruction fields. By tracking creator profiles, developers can study how different teams organize their custom instructions, compare version updates, and fork public configs for specialized enterprise use cases.
SkillMD operates as a high-performance index running on a fast Go backend and a highly responsive Astro SSR frontend. All search queries execute in milliseconds, featuring smart debouncing to prevent multiple API requests while keeping user data secure. Join our community of developers to standardize your AI agent instructions and optimize your LLM prompting workflows today.
Frequently Asked Questions
A practical guide to agent skills: what they are, how to inspect them, and how SkillMD helps you explore the ecosystem.