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...
question
by dancinlabQuick side-question alias — `/question <text>` (and short `/q <text>`) is a short alias for the built-in `/btw <text>` (ask a quick side question without interrupting the current task). Triggers — "/question", "/q", "quick question", "잠깐", "사이드 질문".
quota
by dancinlab/quota:quota <verb> — Claude account 5h/7d usage limits + multi-account registry + live credential swap + per-account nicknames. Single-dispatch picker entry. Verbs — status (default · 6-col 한글 헤더 표 in chat) · list (registered accounts) · add [<nick>] · nick <ref> [<nick>] (set/clear nickname) · switch <ref> · remove <ref> · refresh [<ref>] · help. <ref> = nickname · email · numeric index. Triggers — "quota", "/quota:quota", "쿼터", "한도 보여줘", "session limit 얼마", "주 한도 남았어", "5h 한도", "7d 한도", "claude usage", "내 사용량", "계정 등록", "계정 전환", "계정 바꿔", "다른 계정으로", "claude1으로", "닉네임", "별명".
watch
by dancinlabWatch a video (URL or local path). Downloads with yt-dlp, extracts auto-scaled frames with ffmpeg, pulls the transcript from captions (or Whisper API fallback), and hands the result to Claude so it can answer questions about what's in the video.
all-bg-go
by dancinlabFan out the prior turn's branches in parallel — reactive single fan-out of what the immediately-preceding turn offered. `/abg` is the 3-char alias. Triggers — "all bg go", "all bg", "abg", "다 병렬", "전부 병렬 발사", "전부 다 가자", "fan it all out", "do them all in parallel". For a self-generating loop use /cycle.
all-fg-go
by dancinlabRun the prior turn's branches one at a time in the foreground — reactive single fan-out of what the immediately-preceding turn offered, executed SEQUENTIALLY (no parallelism, no background). `/afg` is the 3-char alias. Triggers — "all fg go", "all fg", "afg", "다 순차", "전부 순차 실행", "하나씩 다 해", "run them all in order", "do them all sequentially". The foreground-sequential sibling of all-bg-go. For a self-generating loop use /cycle-fg.
atlas
by dancinlabRun `hexa atlas` — the atlas SSOT surface (SSOT = compiler/atlas/embedded.gen.hexa, served at runtime by static_atlas via TEXT-parse). Read verbs hash · stats [--audit] · lookup <K> <id> · dump [K]. Write verb register --from-verify <fn> <n> <v> · register --from-drill --seed "<text>" (folds the verified node DIRECTLY into embedded.gen.hexa — no rebuild, no staging shard). Export verb export [--out PATH] (portable .n6 artifact). Triggers — "atlas lookup", "atlas stats", "아틀라스 조회", "아틀라스 등록", "atlas absorb", "register from drill".
brainstorm
by dancinlabIterative brainstorming — given a seed, generate ideas in rounds and keep going until depletion (breadth over selection). Triggers — "brainstorm <X>", "ideate on <X>", "exhaust ideas on <X>", "쥐어짜봐", "모든 각도 생각해봐", "ideate until empty", "exhaust the well".
bypass
by dancinlabAnti-punt — DEFAULT BEHAVIOR (auto-fires, not opt-in). Before any move that hands control back to the user, run the self-check; if local + reversible + non-destructive + no human-only input, just execute. Cross-project enforcement in commons.tape g18. Explicit triggers — "bypass", "그냥 해", "그냥 진행해", "do it yourself", "just run it".
check
by dancinlabIn-flight progress dashboard for THIS session ONLY — report ONLY the progress of work currently running in this session: TaskList state · active monitors · background shell/agent jobs (ps subtree of this session's claude proc) + their fresh output (scoped by CLAUDE_CODE_SESSION_ID). Triggers — "/check", "상태", "체크", "현황", "진행상황", "status check", "dashboard".
cloud
by dancinlabWrap `hexa cloud` for runpod / vast.ai GPU dispatch — structured argv, never raw ssh/scp/runpodctl/vastai/REST (commons g8 · enforced by cloud-guard hook). Subverbs preflight · run · nohup · fire · poll · tail · copy-to · copy-from · pods · dispatch. tail = live remote-log stream → Monitor bridge (commons g57). `pods`/`dispatch` manage the per-project active pod/job work manifest at cwd's `./pods.json` (update-form — sibling/distinct from global `~/.hexa-cloud/pods.jsonl` orphan registry; this is the operator's "what is running in THIS project right now" view per repo). Triggers — "runpod dispatch", "GPU pod에 돌려", "학습 cloud에 던져", "pod 결과 가져와", "로그 실시간 보기", "OOM 사전체크", "preflight", "pods", "cloud pods", "cloud dispatch", "활성 pod", "지금 뭐 돌고있어", "작업 매니페스트", "verdict 갱신", "pod tree".
cycle
by dancinlabAutonomous work-loop driver — enumerate the active domain's open milestones (commons @D g58) → plan table → fan out one background Agent per item → auto-continue to DEPLETION. The whole family (bare /cycle · /cycle-full · /cycle-fg-loop · /cycle-bg-loop · /cycle-full-loop · /cycle-all) drains the domain's `## deferred` backlog to the bottom by self-continuing (ScheduleWakeup) each round until open milestones = 0 AND deferred empty AND no other signal. /cycle-all is the "run everything" variant — no per-round cap + no recommend/select gate, fan out EVERY PROCEED item this round (resource-serialized). Triggers — "/cycle", "/cycle-all", "사이클", "계속 진행", "다음 라운드 진행", "keep cycling", "march on", "next round", "전부 진행", "추천 말고 전부". Distinct from all-bg-go (reactive); /cycle self-generates each round, scoped to the session's active domain.
deck
by dancinlab/deck <domain> <slug> '<spec-json>' — 🍞 도메인 input deck 빵틀 generator. Thin wrapper around `hexa-lang stdlib/deck/gen.hexa`. RTSC QE el-ph first impl (4 files emit — vc-relax.in · scf.in · ph.in · run.sh; spec = anchor · press · q_grid · sscha? · soc?; PSL 1.0.0 pseudo auto-download branch on `soc`). Output dir = exports/<domain>/decks/<slug>/. chem/chip/bio/nuclear/material stub (TODO). Generic dispatch (d4) — domain routing by manifest, no per-domain hardcoding. Preflight estimate reported after emit. d16 pool ubu-1 free dry-run hint before cost-bearing rent. Triggers — '/deck', '빵틀', 'deck-gen', 'input deck 생성', 'bake deck', '빵 굽기', 'deck 만들어', 'cooking deck'.
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.