⚡ Agent Skills Directory
Installable AI agent skills for every profession — 47 tracks, one command
Software · AI/ML · Security · Data · Product · Design · Leadership · HR · Healthcare · Biotech · Science · Finance · Legal · Real Estate · Insurance · Energy · Supply Chain · Hospitality · F&B · Retail · Marketing · Sales · Support · Education · Sports · Media · Government · Nonprofits · Startups · Freelance
580 curated skills and counting. A library of production-grade Agent Skills (SKILL.md procedures) that give coding agents — OpenCode, Claude Code, Cursor, Codex, Copilot, Gemini CLI and 75+ more — the procedural knowledge of an entire senior team across dozens of professions.
Every skill contains: concrete numbered steps, common pitfalls, success signals/KPIs, and copy-paste commands. Authored from real workflows — no filler, no near-duplicate stamping, no AI slop.
🚀 Quick Start
npx skills add aniruddhaadak80/skills
Install one skill to one agent:
npx skills add aniruddhaadak80/skills --skill rag-pipelines-chunking-strategy -a opencode
Your agent discovers installed skills automatically and activates them on demand via progressive disclosure — zero context cost until relevant.
📚 The Library
Software Engineering
| Track | Sample skills |
|---|
| 🤖 AI Engineering | RAG chunking & hybrid search · agent guardrails · prompt-injection defense · LLM cost controls |
| 🎨 Frontend | RSC boundaries · Core Web Vitals rescue · design tokens · animation craft · a11y audits |
| ⚙️ Backend | REST/gRPC design · idempotency keys · zero-downtime migrations · cache layering |
| 🛢️ Data Engineering | Pipeline idempotency · schema contracts · warehouse DQ tests (Snowflake/BigQuery/Databricks) |
| ☁️ DevOps & Cloud | Production Dockerfiles · deploy strategies · Terraform modules · Vercel token deploys |
| 🔐 Cybersecurity | Threat modeling · secrets hygiene · IAM least-privilege · breach first-hour playbook |
| 🛡️ Cyber Defense & SOC | Alert triage · threat hunting · detection engineering (Splunk/Sentinel) · purple-team validation |
| 🧰 Engineering Essentials | Systematic debugging · code review · refactoring · profiling — tuned per stack (Node/Python/Go/Rust/Java/.NET) |
| 🧪 QA & Testing | Test pyramids · flaky-test elimination · maintainable E2E (Playwright/Cypress) |
Data, Science & Research
| Track | Sample skills |
|---|
| 📊 Data Science | Dataset health checks · trustworthy A/B tests · leakage prevention · metric contracts |
| 🧮 ML Research | Paper reproduction protocol · ablation design · benchmark integrity · datasheets |
| 🧬 Biotech & Pharma | Sequencing QC · variant calling SOPs · sample lineage · GxP documentation · clinical data |
| 🔬 Scientific Research | Power analysis · control design · computational reproducibility · grant narratives |
Product & Business
| Track | Sample skills |
|---|
| 📋 Product Management | Customer interviews · outcome roadmaps · one-page PRDs · launch checklists |
| 🚀 Startup & Founders | MVP scope-cutting · founder narrative decks · first growth loops |
| 💼 Freelance & Business | Proposals that close · value pricing · scope-change control |
| 📣 Marketing & Growth | Messaging hierarchies · content engines · lifecycle email flows |
| 🛍️ E-commerce & Retail | PDP conversion audits · checkout friction · returns-as-signal |
| 🤝 Sales & Customer Success | Discovery frameworks · pipeline hygiene · churn-risk signals · QBRs that expand |
| 💬 Customer Support | De-escalation writing · diagnostic trees · KB deflection loops · support QA |
| ☂️ Insurance Operations | FNOL-to-resolution claims · ethical fraud triage · audit-proof underwriting |
| 🏘️ Real Estate | Property underwriting · rigorous comps · renovation ROI · fiduciary representation |
People, Health & Society
| Track | Sample skills |
|---|
| 🧭 Engineering Leadership | Trust-building 1:1s · performance calibration · tech-debt portfolios · structured hiring |
| 👥 People & HR | Inclusive JDs · bias-guarded screening · onboarding/offboarding lifecycle · footprint audits |
| 🏥 Healthcare & Clinical | ISBAR handoffs · med reconciliation · PDSA quality cycles · teach-back education |
| 🎓 Education & Teaching | Assessable objectives · attention-rhythm lessons · course completion mechanics · fair rubrics |
| 🌍 Nonprofit & Impact | Theory of change · impact measurement · donor stewardship · board effectiveness |
| 🏛️ Public Sector & GovTech | Service-standard audits · lock-in-proof procurement · transparency-ready records |
Operations & Specialties
| Track | Sample skills |
|---|
| ⚖️ Legal & Compliance | GDPR data mapping · DSAR flows · MSA redline triage · IP assignment hygiene |
| ✍️ Technical Writing | Diátaxis audits · docs-as-code · endpoint docs · changelogs users read |
| 🧠 UX Research & Design | Usability testing · bias-free synthesis · journey mapping · design critique rituals |
| 🚚 Supply Chain & Logistics | Demand baselines · safety-stock policy · disruption playbooks · freight audits |
| 🍽️ Food & Beverage Ops | Weighted recipe specs · HACCP log discipline · menu engineering · allergen chain-of-custody |
| 🍸 Beverage & Bar Program | Cocktail spec cards · weight-based inventory control · responsible service flow |
| 🏨 Hospitality & Travel | Service recovery · itinerary pacing · dynamic pricing ethics · review response systems |
| 🏆 Sports & Performance | Periodization · load monitoring · criteria-based return-to-play · effective cueing |
| 🎬 Media & Video Production | Pre-pro checklists · interview craft · edit pacing discipline · publish packaging |
| ✈️ Aviation & Aerospace Ops | Checklist culture · go/no-go gates · just-culture reporting · maintenance planning |
| 🐾 Veterinary Medicine | Urgency triage · cost-consent clarity · controlled-drug logs · zoonotic safety |
| 🏗️ Construction & Built Environment | Toolbox talks · schedule recovery · inspection test plans · estimate discipline |
| 🌾 Agriculture & AgTech | Rotation design · input stewardship · yield-map analysis · herd records |
| 🌐 Localization & i18n | String externalization · RTL support · LQA cycles · TMS workflows |
| 🎯 Career & Job Search | Truth-effective resumes · STAR story banks · principled offer negotiation |
| 🎮 Game & Interactive Dev | Game-feel tuning · crash-safe save systems · frame budgets · live-ops pipelines |
| 🔋 Energy & Renewables | Solar feasibility screens · P50/P90 model integrity · commissioning rigor · storage cycling |
Plus 🗺️ Journey Playbooks — multi-skill guided sequences like "Launch a RAG feature end to end", "New engineering manager: first 90 days", "Stand up a privacy program baseline", and "Screen a solar project like a developer".
🌐 Live Directory
Browse, search (Ctrl+K), filter, and copy install commands at the companion web app — Next.js, fully static, animated, RSS feed included:
npm install && npm run dev # local development
npm run gen # regenerate SKILL.md files from src/data/
node scripts/verify-routes.mjs # verify every route exists in build output
🏗️ How It Works
src/data/tracks/*.mjs ← structured knowledge base (source of truth)
│
▼ node scripts/generate.mjs
skills/<slug>/SKILL.md ← Agent Skills format, auto-generated + indexed
│
▼ npx skills add aniruddhaadak80/skills
Your agent ← discovers + activates skills on demand
Each skill follows the open spec:
---
name: my-skill
description: What it does + when the agent should activate it
---
# Steps, pitfalls, success signals, commands…
The generator also emits public/search-index.json (powers the site's Ctrl+K palette) and validates slug uniqueness across the whole library.
📈 Scaling Philosophy
The architecture is built to reach six figures of skills without quality collapse:
- Expansion packs — new domains per track; each procedure = an installable skill
- Stack/tool variants — universal procedures multiply with genuinely stack-specific steps
- Playbook compositions — new journeys assemble existing skills into guided sequences
- Community contributions — PRs adding procedures auto-generate installable skills + site pages
Honest note from the author: this repo ships only skills that pass a quality bar — real steps, real pitfalls, no template spam. Volume follows quality, never the reverse.
🤝 Contributing
Read CONTRIBUTING.md — it covers the quality bar, full schema reference, and local workflow. Short version:
- Pick a track you have real production experience in
- Add a procedure object to
src/data/tracks/<track>.mjs following the schema
- Run
npm run gen + node scripts/validate-data.mjs — your SKILL.md, website page, and index entry appear automatically
- Open a PR; automated review checks everything (data QA, generated-file sync, route integrity) and labels it instantly
Every merged PR earns you a place on the auto-updated contributors wall below.
🌟 Contributors (1)
Thank you to everyone who has helped build the Agent Skills Directory!
MIT © Aniruddha Adak · 350+ merged OSS PRs · [email protected]
