🧰
overdrive
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. overdrive from pbakaus/impeccable: ──────────── ⚡ OVERDRIVE ───────────── 》》》 Entering overdrive mode...
skills.shagent skillpbakaus
80
Learn AI 스킬
AI 도구와 기술 마스터하기
161 ai 스킬
📣
marketing-ideas
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. marketing-ideas from coreyhaines31/marketingskills: Brainstorm and prioritize marketing strategies tailored to your SaaS stage, budget, and goals.
skills.shagent skillcoreyhaines31
90
Learn 💻
typeset
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. typeset from pbakaus/impeccable: Systematically assess and refine typography to eliminate generic defaults and establish clear hierarchy, readability, and brand personality.
skills.shagent skillpbakaus
60
Learn 🧩
writing-skills
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. writing-skills from obra/superpowers: Test-driven documentation for creating reusable agent techniques, patterns, and reference guides.
skills.shagent skillobra
90
Learn 📣
programmatic-seo
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. programmatic-seo from coreyhaines31/marketingskills: Build SEO-optimized pages at scale using templates, data, and proven playbook patterns.
skills.shagent skillcoreyhaines31
50
Learn 📣
content-strategy
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. content-strategy from coreyhaines31/marketingskills: Plan searchable and shareable content that drives traffic, builds authority, and generates leads.
skills.shagent skillcoreyhaines31
40
Learn 🧪
webapp-testing
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. webapp-testing from anthropics/skills: Native Python Playwright scripts for testing local web applications with server lifecycle management.
skills.shagent skillanthropics
80
Learn 🎬
codex-pet
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. codex-pet from agentspace-so/runcomfy-agent-skills: codex-pet packages reusable procedural knowledge for agents.
skills.shagent skillagentspace-so
90
Learn 🧩
receiving-code-review
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. receiving-code-review from obra/superpowers: receiving-code-review packages reusable procedural knowledge for agents.
skills.shagent skillobra
70
Learn 🧰
lark-approval
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. lark-approval from open.feishu.cn: lark-approval packages reusable procedural knowledge for agents.
skills.shagent skillopen.feishu.cn
90
Learn 🧪
verification-before-completion
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. verification-before-completion from obra/superpowers: Enforce verification commands before claiming work completion, fixes, or test passes—no success assertions without fresh evidence.
skills.shagent skillobra
80
Learn 🧰
subagent-driven-development
Intermediate 20-45 minutes
skills.sh 생태계의 재사용 가능한 Agent Skill입니다. subagent-driven-development from obra/superpowers: Dispatch fresh subagents per task with two-stage review (spec compliance, then code quality) in the current session.
skills.shagent skillobra
70
Learn 
Skill engineering
AI skills are becoming the reusable capability layer between prompts and agents.
A skill is not just a long prompt. It is a packaged procedure: activation metadata, task instructions, optional scripts, reference files, examples, and validation steps. Good skill pages should explain what the skill does, when it activates, what resources it loads, and how users can verify the result.
What changed
Clarify capability boundaries before adoption.
Packaged procedure
Skills store repeatable work, not background trivia
Claude describes skills as directories with instructions, scripts, and resources, with a SKILL.md file defining activation and behavior. That makes skills suitable for team workflows such as SEO audits, spreadsheet analysis, slide generation, brand writing, and code review.
Progressive disclosure
Skills help agents load less context upfront
Claude’s documentation describes a three-step pattern: lightweight metadata at startup, full instructions only when the task matches, and extra resources only when needed. This directly addresses context-window overload.
Runtime integration
Skills are moving into agent sandboxes
OpenAI’s Agents SDK includes a skills capability that mounts skills into an auto-discovery root inside a sandbox and supports lazy loading. In practice, that turns skills into portable operating procedures an agent can discover during work.
Popular starting points
Build AI skills for work, learning, and career growth
Explore AI skills by learning goal: find an AI skills course, learn AI skills for free, compare structured programs, understand AI skills in demand, and map skills to AI engineer requirements.
ai skills
Core learning topicA practical starting point for prompt engineering, AI tools, agent workflows, data skills, evaluation, and applied workplace AI literacy.
how to learn ai skills
Learning pathBest for users who need a staged learning path from beginner foundations to practical projects and advanced agent workflows.
ai skills in demand
Career and commercial valueUseful for career-focused readers comparing workplace AI skills, AI engineer skills required, and portfolio projects that prove ability.
Explore AI skills learning paths
Filter by the job to be done.
The page should combine learning, certification, role readiness, and agent-skill workflows instead of treating skills as generic tutorials.
Learning path intent
These queries need a staged path: foundations, prompt skills, AI tools, agent workflows, data handling, evaluation, and portfolio projects.
how to learn ai skillsai skills to learnhow to learn ai skills for freeai skills course
Program and certificate intent
These searches point to structured programs. The page can cover how to compare AI skills passport, Microsoft AI Skills Fest, and similar learning programs without overclaiming affiliation.
ai skills passportey ai skills passportmicrosoft ai skills fest 2025ai skills fest
Career intent
These users care about employability. Add sections about AI engineer skills required, workplace AI literacy, and evidence of skill through projects.
ai skills in demandai engineer skills requiredai apprenticeship skills england
Skills vs prompts vs agents
Compare by capability, cost, and risk.
These three layers should work together. A prompt instructs one task, a skill packages repeatable expertise, and an agent executes across tools and time.
Prompt
- A single instruction or template
- Fast to copy and edit
- Weak at storing procedures, assets, and scripts
Skill
- A reusable task package
- Can include instructions, scripts, examples, and references
- Best for repeatable expert workflows
Agent
- A system that plans and acts
- Uses tools, memory, and skills
- Needs governance, observability, and permissions
What makes a skill worth learning
Fit the tool into a real workflow.
It solves a recurring task with clear completion criteria.
It contains examples or scripts that reduce manual work, not just advice.
It defines when to activate and when not to activate.
It includes a quality check: test, render, audit, cite, compare, or validate before final output.
Related AI skills topics
Use these topics to move from basic AI skills into courses, free learning paths, career skills, engineering requirements, and agent skill workflows.
ai skillsai skills coursehow to learn ai skillsai skills to learnhow to learn ai skills for freeai skills in demandai engineer skills requiredai agent skillsai skills hub
Skill FAQ
Common questions and selection answers.
What should a skill page explain?
Explain the task it handles, the trigger conditions, the workflow, required tools or files, expected output, quality checks, and common failure cases. That is more useful than listing broad benefits.
How are skills different from projects or custom instructions?
Projects and custom instructions are broad background context. Skills are task-specific procedures that load dynamically, which makes them better for specialized work that should not always be in the context window.
Which skills are high-value for AI teams?
High-value skills encode work that happens often and has a review standard: SEO content audits, research briefs, spreadsheet modeling, slide creation, code review, data cleaning, customer support QA, and brand-compliant writing.