cosmicstack-labs/mercury-agent-skills

최근 기록된 소스 활동: 2026년 8월 18일 09:29

SkillsMP 카탈로그 업데이트: 2026년 10월 29일 18:30

수집된 skills: 137

GitHub 스타: 482
GitHub 포크: 65

AI 어시스턴트로 설치

이 프롬프트를 복사해 사용 중인 AI 어시스턴트에 입력하세요.

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/cosmicstack-labs/mercury-agent-skills.

저장소 README

README 원본 파일: README.md

README 원본 보기

README 전체 펼치기

# Mercury Skills <img src="assets/mercury-agent-skills-card.png" alt="Mercury Skills" width="75%" style="max-width: 700px; height: auto;"> <a href="https://skills.mercuryagent.sh">Browse the registry</a> • <a href="https://mercuryagent.sh">Mercury Agent</a> • <a href="./CATALOG.md">Catalog</a> • <a href="./CONTRIBUTING.md">Contribute</a> **An open-source library of `SKILL.md` playbooks for AI agents.** Browse and install on [skills.mercuryagent.sh](https://skills.mercuryagent.sh), or pull them directly from the [Mercury Agent](https://mercuryagent.sh) CLI. **132 skills across 23 categories** — hand-curated, production-ready, and compatible with every major agent: Mercury, Claude Code, Codex CLI, OpenClaw, Hermes, Cursor, and Gemini CLI. --- ## Quick Start ### From the Mercury CLI (recommended) ```bash # List what's available mercury skills search "react patterns" # Install one skill mercury skills install frontend/react-patterns # Install several at once mercury skills install ai-ml/prompt-engineering devops/docker security/threat-modeling ``` Installed skills land in `~/.mercury/skills/<category>/<slug>/SKILL.md` and are picked up automatically by your agent. ### From the web Open [**skills.mercuryagent.sh**](https://skills.mercuryagent.sh), pick a skill, copy the install line for your agent — every skill page ships ready-made install steps for Mercury, Claude Code, Codex, OpenClaw, Hermes, and "any other agent that understands SKILL.md". ### Clone the whole library ```bash git clone https://github.com/cosmicstack-labs/mercury-agent-skills.git ``` Drop individual skills into your agent's skills directory: | Agent | Skills directory | |---|---| | **Mercury** | `~/.mercury/skills/` | | **Claude Code** | `.claude/skills/` | | **Codex CLI** | `.codex/skills/` | | **OpenClaw** | `.openclaw/skills/` | | **Hermes** | `.hermes/skills/` | | **Cursor** | `.cursor/skills/` | | **Gemini CLI** | `.gemini/skills/` | | Any other | Wherever your agent reads `SKILL.md` from | --- ## Why Mercury Skills? | | | |---|---| | **Curated, not crowded** | Every skill is hand-written against real workflows — no AI slop, no duplicates. | | **Universal format** | One `SKILL.md` works in any agent that supports the standard. No per-agent forks. | | **Discoverable** | Full-text search, categories, tags, leaderboard, and live JSON feed on [skills.mercuryagent.sh](https://skills.mercuryagent.sh). | | **Installable** | One CLI command, no manual file copying. Atomic writes, version pinning, rollback on failure. | | **Open** | MIT licensed. Contributions welcome via PR — see [CONTRIBUTING.md](./CONTRIBUTING.md). | --- ## Browse on the Web [**skills.mercuryagent.sh**](https://skills.mercuryagent.sh) is the canonical browse-and-install surface for the library. It serves: - A searchable, categorized index of every skill in this repo - A trending leaderboard ranked by likes and installs - Per-skill detail pages with the full `SKILL.md`, install steps for every supported agent, and a copy-paste-ready install command - A bookmarking system so you can save skills for later (local, no account required) - Social share cards with per-skill OpenGraph imagery It also exposes a public JSON feed at [`/api/feed.json`](https://skills.mercuryagent.sh/api/feed.json) and per-skill JSON at `/api/skills/<category>/<slug>` for tooling builders. --- ## Categories | Category | Skills | What's inside | |---|---|---| | [Development](./categories/development/) | 16 | Clean code, code review, debugging, testing, ADRs, documentation, refactoring, dependency hygiene | | [AI & ML](./categories/ai-ml/) | 11 | Prompt engineering, agent health, memory, delegation, handoffs, token budgets, error recovery, audit logging | | [Backend](./categories/backend/) | 9 | APIs, Node.js, Python, database design, auth, serverless, microservices, caching, message queues | | [Shop & Restaurant](./categories/shop-restaurant/) | 10 | Inventory, menu engineering, scheduling, reviews, daily pulse, table management, pricing, social, Zomato ordering, Amazon assistant (search/cart/orders/insights) | | [Frontend](./categories/frontend/) | 8 | React, Next.js, Tailwind, state management, testing, performance, responsive design, component systems | | [Creative & Personal Development](./categories/creative-personal-development/) | 8 | Storytelling, decision frameworks, standups, notes, repurposing, branding, validation, time blocking | | [DevOps](./categories/devops/) | 7 | Docker, CI/CD, Kubernetes, Terraform, monitoring, SRE, GitOps | | [Career](./categories/career/) | 5 | Resume writing, interview prep, career planning, LinkedIn, salary negotiation | | [Education & Learning](./categories/education-learning/) | 5 | Curriculum design, learning science, teaching methods, assessment, micro-learning | | [Finance & Legal](./categories/finance-legal/) | 5 | Financial analysis, budgeting, contracts, privacy compliance, risk management | | [Health & Wellness](./categories/health-wellness/) | 5 | Fitness, nutrition, mental health, sleep, habits | | [Mobile](./categories/mobile/) | 5 | iOS, Android, React Native, performance, App Store optimization | | [Testing & QA](./categories/testing-qa/) | 5 | Test strategy, E2E, performance testing, API testing, accessibility testing | | [Automation](./categories/automation/) | 11 | Screenshots, workflows, shell scripting, web scraping, X/Twitter automation, twitter-account-manager (research), test automation, data sync, deployment automation, RPA, daily briefing | | [Media Download](./categories/media-download/) | 6 | yt-dlp wrappers, audio extraction, batch downloading, platform-specific patterns, GitHub repo promo videos, GitHub repo scrolling tours | | [PDF Generation](./categories/pdf-generation/) | 5 | Invoices, reports, branded documents, dynamic templating, markdown→PDF typesetting (any2pdf) | | [Presentation](./categories/presentation/) | 4 | Slide decks, narrative structure, stage delivery, visual hierarchy | | [Marketing](./categories/marketing/) | 3 | SEO, content strategy, distribution | | [Business](./categories/business/) | 2 | Negotiation, operations | | [Design](./categories/design/) | 2 | UI systems, accessibility | | [Product](./categories/product/) | 2 | Strategy, discovery | | [Security](./categories/security/) | 2 | Audit, secure coding | | [Data](./categories/data/) | 1 | Data engineering fundamentals | See [CATALOG.md](./CATALOG.md) for the full skill-by-skill index. --- ## Skill Structure Every skill is a single `SKILL.md` file with YAML frontmatter and a markdown body: ```yaml --- name: skill-name description: 'What this skill does and when to use it' metadata: author: cosmicstack-labs version: 1.0.0 category: development tags: [clean-code, refactoring, best-practices] --- # Skill Name Full instructions, frameworks, scoring rubrics, and actionable guidance. ``` The format is intentionally minimal so any agent runtime can parse it without custom tooling. See [CONTRIBUTING.md](./CONTRIBUTING.md) for the authoring standard. --- ## Try Mercury Agent These skills are built for [**Mercury Agent**](https://mercuryagent.sh) — the agent runtime that consumes them natively. - **Second Brain** — persistent memory that learns from every session - **Skill system** — load, version, and update skills from this registry directly - **Permission guardrails** — safe by design, auditable by default - **Token budgets** — keep AI spend predictable - **Multi-channel** — CLI, Telegram, and web; same agent, same memory ```bash # Install Mercury Agent npm install -g @cosmicstack/mercury-agent # Pull a skill into your agent mercury skills install ai-ml/prompt-engineering # Browse what you have mercury skills list ``` [**Get started with Mercury →**](https://mercuryagent.sh) --- ## Ecosystem | Project | What it is | |---|---| | [**Mercury Agent**](https://mercuryagent.sh) | The agent runtime — CLI, Telegram, web — that loads these skills | | [**Mercury Skills (this repo)**](https://github.com/cosmicstack-labs/mercury-agent-skills) | The curated `SKILL.md` library | | [**skills.mercuryagent.sh**](https://skills.mercuryagent.sh) | The browse-and-install web platform powered by this repo | Built and maintained by [Cosmic Stack](https://cosmicstack.org). --- ## Contributing We accept new skills, fixes, and refinements via pull request. - **Author a skill** — see the standard in [CONTRIBUTING.md](./CONTRIBUTING.md) - **Fix or improve an existing skill** — open a PR with a clear changelog note - **Request a category** — open an issue with the proposed scope Every merged contribution lands on [skills.mercuryagent.sh](https://skills.mercuryagent.sh) within the hour via an automated sync. --- ## License MIT — see [LICENSE](./LICENSE). Built by <a href="https://cosmicstack.org">Cosmic Stack</a> · Powered by <a href="https://mercuryagent.sh">Mercury Agent</a>

저장소 README를 불러오는 중…

이 저장소의 skills

소프트웨어 개발자 시장조사 분석가·마케팅 전문가 재무 및 투자 분석가 소프트웨어 품질 보증 분석가·테스터

직업 카테고리 23개 · 100% 분류됨

소프트웨어 개발자

Async control flow, error handling, module design, streams, and production hardening

원문 언어: 영어

2026-05-10

수집된 skill 137개 중 40개를 표시합니다.