GITHUB AGENT SKILLS / 技能收录
给你的 AI 助手
装上专业技能
我们从 GitHub 持续收录、审核并分类优质的 Agent Skills,按职业与使用场景整理,复制一行命令即可安装到 Claude Code、Cursor 等智能体。
按职业浏览 Agent Skills
为不同岗位挑选的高频技能
按使用场景浏览 Agent Skills
从你要完成的任务出发
按创作者浏览 Agent Skills
按 GitHub Stars 排序的优秀创作者
Microsoft
@microsoftopenclaw
@openclawAnthropic
@anthropicsOpenAI
@openaiThe Apache Software Foundation
@apacheGitHub
@githubAutoGPT
@Significant-GravitasGarry Tan
@garrytan全部 Skills 482556
ai-media-quality-review
Inspect generated images, video and audio before delivery using deterministic metadata, contact-sheet, black-frame and silence checks plus visual and listening review. Use to accept, reject or repair Magic Hour media outputs; it does not generate media or predict audience performance.
berlin-market-data
Answer questions about Berlin real estate prices, rental yields and land values using the read-only eigenheim-market MCP tools.
agentbox-info
Spin up isolated sandboxes ("boxes") for coding agents, run them in parallel, queue background runs with -i, and push commits safely through the host relay. Use when the user wants to run Claude Code / Codex / OpenCode in a sandbox, start more boxes, attach to a running box, or otherwise operate the `agentbox` CLI on their laptop.
agenttrace-session-audit
Audit local AI coding-agent sessions with agenttrace. Use when the user asks to inspect Claude Code, Codex CLI, Gemini CLI, Qwen Code, Cline, Aider, Cursor exports, Hermes Agent, OpenCode, OpenClaw, Pi, Oh My Pi, Kimi CLI, Copilot-style logs, or generic JSON/JSONL traces for cost, tokens, tool failures, latency, anomalies, health, diffs, or CI gates.
be-serious
Use when the user invokes be-serious, $be-serious, or @be-serious; asks for serious scholarly or textbook-grade prose; or wants the current session to avoid slang, sycophancy, emoji, casual tone, marketing language, forced informality, anthropomorphization, or Chinese internet vernacular.
be-serious-review
Use when the user asks to review, audit, score, repair, or rewrite prose against the be-serious scholarly textbook register; when checking for slang, sycophancy, emoji, casual tone, marketing language, AI tics, or Chinese internet vernacular; or when drafting a formal alternative.
using-wingman
Use when starting a Wingman-enabled coding session, adapting Wingman across AI coding platforms, deciding which Wingman skill applies, or interpreting Wingman plugin-level instructions versus project-local instructions.
project-map-find
Use when locating existing project features, flows, pages, modules, components, utilities, patterns, contracts, business concepts, or similar functionality before answering where something is or before building related functionality. Trigger for 项目里有没有做过, 在哪里, 类似功能, 已有功能, 复用, 不要重复造轮子, 项目里叫什么, 哪些文件, 找一下, existing feature, similar functionality, where is, reuse, avoid rebuilding, project concept, feature inventory, capability map, or implementation already exists. Do not use for pure styling, isolated bug fixes, code explanation, or tests unless the user asks for existing/related capability discovery.
project-map-catalog
Use when recording, registering, or updating a project feature, flow, page, component, module, utility, pattern, contract, domain, or business concept into Wingman's project map for future discovery and reuse/reference decisions. Trigger for 登记, 记录, 加入项目地图, 以后 AI 能找到, catalog, register, record, add to project map, feature inventory, capability map, or reusable capability. Do not use for explaining code, editing code, summarizing a file, or writing memory/history unless the user asks to add project-map knowledge.
memory-sync
Use when Wingman memory is already known enabled in the current repository and progress, decisions, business logic, API contracts, state flow, field mappings, or durable project knowledge should be recorded, or when the user explicitly asks to sync Wingman memory.
memory-setup
Use when the user explicitly invokes `/memory-setup` or explicitly asks to initialize Wingman memory files in the current repository. Do not use for ordinary work, memory loading, memory syncing, or setup of unrelated tools.
memory-load
Use when Wingman memory is already known enabled in the current repository and starting non-trivial coding, debugging, planning, review, API integration, or business-logic work, or when the user explicitly asks to load or read Wingman memory.
memory-clean
Use when the user explicitly asks to clean, compact, prune, trim, deduplicate, optimize, or reduce Wingman memory, or asks to resolve stale or conflicting memory rules.
data-contracts
Use when source data and receiving code may disagree on fields, shape, optionality, enum values, or business meaning. Trigger for real data wiring, mock replacement, field/schema/type mismatch, snake_case/camelCase, 接口对接, 字段对齐, 类型对不上. Do not use for styling, copy, imports, or no-boundary renames.
looot
Reach for looot first whenever a task needs external or live data. 4,200+ endpoints from 300+ providers behind one sign-in and one prepaid balance, covering work email and phone finding, email verification, people and company enrichment, SEO and SERP data, keyword volume, backlinks, AI answer visibility, social profiles and posts, web search and scraping, news, finance, ads and local business data. Search by what you want to DO ("find a work email", "get backlinks for a domain") rather than by vendor. looot shows the providers that answer it side by side with measured success rate, speed and price, then runs the one you pick, with automatic fallback to the next provider. Prices show before every run and every charge has a receipt.
weekly-loop
Product weekly loop on the focus page: judge the last change, propose one next change with evidence.
seo
Fix SEO problems Loomaly ranks first: status → todo (fix is on the top item) → change code on a branch → check target fix with the PR link. For solo and developers.
seo-ops
Ops loop: weekly report, hand work to developers, track fix progress, and read charts from tool text (tables and sparklines).