Antigravity Awesome Skills — Open Source Project Introduction

Repository: https://github.com/sickn33/antigravity-awesome-skills

1. Project Overview

Antigravity Awesome Skills is a curated, battle-tested library of 1,239+ universal agentic skills (structured SKILL.md markdown files) that supercharge AI coding assistants — Claude Code, Gemini CLI, Codex CLI, Cursor, GitHub Copilot, Antigravity IDE, OpenCode, and more — by giving them reusable, domain-specific playbooks for planning, coding, debugging, testing, security review, infrastructure, and beyond.

2. Project Background & Positioning

Why It Exists

AI coding assistants are powerful but inherently general-purpose. They do not know your organization's deployment protocol, the exact syntax of a cloud provider's IaC format, or the nuanced steps of a security audit. Skills solve this gap: small, structured markdown files that teach an AI agent exactly how to perform a specific, repeatable task — consistently and correctly every time.

The project was born out of the need to consolidate scattered, fragmented community skill experiments into a single, quality-controlled, cross-tool-compatible library, including official skills from Anthropic, Google, Vercel, Microsoft, Supabase, and OpenAI.

What Sets It Apart

Dimension	Antigravity Awesome Skills	Typical Prompt Libraries
Scope	1,239+ skills across 9+ domains	Dozens of prompts in one domain
Cross-tool support	Claude Code, Cursor, Gemini CLI, Copilot, OpenCode, Kiro, AdaL	Usually 1 tool
Official sources	Anthropic, Google, Microsoft, Vercel, Supabase, OpenAI	Community only
Installation	One `npx` command	Manual copy-paste
Structure	SKILL.md standard format + role-based bundles + workflows	Flat prompt lists
Web browser	Interactive hosted web app for searching skills	None

3. Feature Category Breakdown

🏛️ Architecture

~30+ skills

architecture — Full system design and component breakdown
c4-context — C4 model diagrams for context and container views
senior-architect — Senior-level architectural review and ADRs
api-design-principles — REST/GraphQL API shape and consistency
ddd-core-domain — Domain-Driven Design core domain modeling

Use these to design scalable, well-documented systems from the ground up.

💼 Business

~60+ skills

copywriting — Marketing and product copy generation
pricing-strategy — SaaS pricing model analysis
seo-audit — On-page SEO review and recommendations
cro — Conversion rate optimization tactics
go-to-market — GTM planning and positioning

Extends AI capabilities into product, marketing, and growth workflows.

🤖 Data & AI

~80+ skills

rag-engineer — Retrieval-Augmented Generation pipeline design
prompt-engineer — Structured prompt crafting and evaluation
langgraph — LangGraph agent graph construction
llm-observability — Tracing and monitoring LLM apps
agent-memory-mcp — Persistent memory patterns for AI agents

Purpose-built for teams building LLM applications, RAG systems, and AI agents.

💻 Development

~300+ skills

typescript-expert — TypeScript patterns, strict typing, utility types
python-patterns — Pythonic idioms, async patterns, testing
react-patterns — Component composition, hooks, performance
go-patterns — Idiomatic Go, concurrency, error handling
rust-patterns — Ownership model, lifetimes, async Rust

The largest category — language and framework mastery for everyday development.

📋 General

~100+ skills

brainstorming — Structured ideation before implementation
doc-coauthoring — Collaborative, structured technical documentation
writing-plans — Project and feature planning documents
create-pr — Clean, well-structured pull request packaging
debugging-strategies — Systematic troubleshooting methodologies

Cross-cutting productivity skills useful across all roles and project types.

☁️ Infrastructure

~150+ skills

docker-expert — Dockerfile optimization, multi-stage builds, compose
aws-serverless — Lambda, API Gateway, SAM, CDK patterns
vercel-deployment — Vercel deployment configurations and edge functions
ci-cd-pipelines — GitHub Actions, GitLab CI, Jenkins workflows
kubernetes-patterns — Helm charts, manifests, resource management

Covers the full DevOps and cloud engineering lifecycle.

🔐 Security

~120+ skills

api-security-best-practices — Auth, input validation, rate limiting
sql-injection-testing — OWASP-aligned SQL injection analysis
vulnerability-scanner — Automated code vulnerability review
security-auditor — End-to-end security audit playbook
owasp-top10 — Systematic OWASP Top 10 review checklist

Sourced from HackTricks, OWASP, and the security research community.

🧪 Testing

~80+ skills

test-driven-development — Red-green-refactor TDD methodology
testing-patterns — Unit, integration, and E2E test design
test-fixing — Diagnosing and repairing failing test suites
go-playwright — Playwright browser automation for Go stacks
qa-workflows — Full QA review and sign-off workflows

Enables disciplined, high-coverage testing across all layers of a stack.

⚙️ Workflow

~50+ skills

workflow-automation — Automated pipeline design and orchestration
inngest — Inngest background job and event-driven function patterns
trigger-dev — Trigger.dev job scheduling and event handling
agent-orchestration — Multi-agent coordination and handoff patterns

Automates repetitive engineering workflows and agent pipelines.

4. Core Highlights

🌍 Universal Compatibility — The universal SKILL.md format works identically across Claude Code, Gemini CLI, Codex CLI, Cursor, Antigravity IDE, Kiro, OpenCode, GitHub Copilot, and AdaL. One library; every tool.
⚡ One-Command Install — A single npx antigravity-awesome-skills command deploys all 1,239+ skills to the correct directory for your chosen tool. No manual file management required.
🏆 Official Sources Included — Bundles official skills directly from Anthropic, Google Gemini, Vercel Labs, Microsoft, Supabase, OpenAI Codex, Apify, and Remotion — not just community contributions.
📦 Role-Based Bundles — Pre-curated starter packs (Web Wizard, Security Engineer, OSS Maintainer, DevOps & Cloud, etc.) help users avoid analysis paralysis across 1,239+ skills.
🗺️ Outcome-Driven Workflows — Step-by-step execution playbooks (e.g., "Ship a SaaS MVP", "Security Audit for a Web App", "Build an AI Agent System") chain skills together into end-to-end processes.
🌐 Interactive Web Browser — A hosted web app at https://sickn33.github.io/antigravity-awesome-skills/ provides full-text search, category filtering, and one-click copy for every skill.

5. Applicable Scenarios (By Role)

👨‍💻 General Developer

Use skill bundles like Essentials + Full-Stack Developer to get structured guidance on feature planning (@brainstorming), API design (@api-design-principles), code review (@lint-and-validate), and pull request packaging (@create-pr) — without leaving your AI coding assistant.

🔧 DevOps / Infrastructure Engineer

The DevOps & Cloud bundle gives you skills for Docker optimization, Kubernetes manifests, serverless deployments, CI/CD pipeline design, and cloud provider IaC — covering AWS, GCP, Azure, and Vercel workflows.

🔐 Security Engineer

The Security Developer bundle draws from OWASP and HackTricks to provide skills for penetration testing methodologies, vulnerability scanning, secure API review, SQL injection analysis, and compliance checklists.

📊 Data / AI Researcher

Use Data & AI category skills to build RAG pipelines (@rag-engineer), craft effective prompts (@prompt-engineer), instrument LLM observability (@llm-observability), and implement LangGraph or LangChain patterns.

🗂️ Project Manager / Tech Lead

Use General skills for structured brainstorming, requirements planning, technical documentation co-authoring, ADR creation, and architecture decision records — bringing consistency to planning workflows.

6. Quick Start

🔍 How to Find Resources

Browse the skill catalog in three ways:

# 1. Interactive web app (recommended)
# https://sickn33.github.io/antigravity-awesome-skills/

# 2. Full skill catalog in the repository
cat CATALOG.md

# 3. Role-based bundles
cat docs/users/bundles.md

🛠 How to Install & Integrate

# Install for your specific tool (examples):
npx antigravity-awesome-skills --claude      # Claude Code → .claude/skills/
npx antigravity-awesome-skills --cursor      # Cursor → .cursor/skills/
npx antigravity-awesome-skills --gemini      # Gemini CLI → .gemini/skills/
npx antigravity-awesome-skills --codex       # Codex CLI → .codex/skills/
npx antigravity-awesome-skills --antigravity # Antigravity IDE

# Verify installation:
test -d ~/.gemini/antigravity/skills && echo "Skills installed successfully"

# Use your first skill:
# In Claude Code:
>> /brainstorming help me plan a SaaS MVP

# In Cursor:
@brainstorming help me plan a SaaS MVP

# In Gemini CLI:
Use brainstorming to help me plan a SaaS MVP

🤝 How to Contribute

# 1. Fork the repository on GitHub
# https://github.com/sickn33/antigravity-awesome-skills/fork

# 2. Clone your fork
git clone https://github.com/<your-username>/antigravity-awesome-skills.git

# 3. Create a new skill following the template
cp docs/contributors/skill-template.md skills/<your-skill-name>/SKILL.md
# Edit SKILL.md following the anatomy guide:
# docs/contributors/skill-anatomy.md

# 4. Validate your skill before opening a PR
npm run validate

# 5. Open a pull request
# See CONTRIBUTING.md for full guidelines

7. Project Structure

antigravity-awesome-skills/
├── skills/                     # The canonical skill library (1,239+ SKILL.md files)
│   ├── brainstorming/
│   │   └── SKILL.md
│   ├── security-auditor/
│   │   └── SKILL.md
│   └── ...
├── docs/
│   ├── users/                  # Getting started, usage guide, bundles, workflows
│   │   ├── getting-started.md
│   │   ├── usage.md
│   │   ├── bundles.md          # Role-based starter packs
│   │   └── workflows.md        # Step-by-step execution playbooks
│   ├── contributors/           # Skill template, anatomy, quality bar
│   ├── maintainers/            # Release, audit, CI drift documentation
│   └── sources/                # Attribution and licensing references
├── apps/web-app/               # Interactive skill browser (Vite app)
├── data/                       # Generated catalog, aliases, bundles, workflows JSON
├── tools/                      # Installer CLI, validators, generators
├── skills_index.json           # Machine-readable skill registry
├── CATALOG.md                  # Human-readable full catalog
├── CONTRIBUTING.md             # Contribution guidelines
├── CHANGELOG.md                # Version history
└── package.json                # npm package (npx installer)

8. Related Ecosystem

Upstream Tools & Platforms

Claude Code (Anthropic) — Primary CLI target for skill invocation
Gemini CLI (Google) — Supported CLI with native skill loading
Codex CLI (OpenAI) — Supported CLI for skill-based prompting
Cursor (Anysphere) — IDE with @skill-name invocation in chat
Kiro (AWS) — IDE/CLI with on-demand skill loading
Antigravity IDE (DeepMind) — Native agent mode with skill orchestration
OpenCode — Open-source CLI with .agents/skills/ path support

Complementary Projects

antigravity-workspace-manager — CLI companion to auto-provision skill subsets across dev environments
VoltAgent/awesome-agent-skills — Source of 61 additional high-quality skills (Sentry, Expo, Hugging Face)
obra/superpowers — The original "Superpowers" project that inspired this collection

9. License

MIT License

✅ Free to use commercially and personally
✅ Allowed to modify and distribute
✅ Allowed to sublicense
✅ Allowed to use in closed-source projects
❌ No warranty provided — use at your own risk
❌ Authors not liable for damages
ℹ️ Attribution required — keep the LICENSE file and copyright notice intact
ℹ️ Individual skills sourced from official repositories (Anthropic, Google, Vercel, etc.) may carry their own upstream licenses — check docs/sources/sources.md for details

10. Frequently Asked Questions

Q: Do I need to install each bundle separately?
A: No. One install (npx antigravity-awesome-skills) gives you all 1,239+ skills. Bundles are just curated reading lists — they require no separate installation.

Q: Why does installation use ~/.gemini/antigravity/skills as the default path even for Claude Code?
A: The default path is Antigravity's global path. Use tool-specific flags (--claude, --cursor, --gemini) or --path to install to the correct location for your tool.

Q: Skills are not being picked up by my AI tool. What's wrong?
A: Check that the skills are in the correct directory for your tool. On Windows, symlinks require Developer Mode — see the Troubleshooting section in the README.

Q: How do I invoke a skill in my AI assistant?
A: Invocation syntax varies by tool. Examples: >> /brainstorming (Claude Code), @brainstorming (Cursor), Use brainstorming to... (Gemini CLI). See the full compatibility table.

Q: Can I contribute my own skills?
A: Yes. Add a new skills/<skill-name>/SKILL.md, follow the skill-anatomy guide, run npm run validate, and open a pull request.

11. Quick Links

Resource	Link
Repository	https://github.com/sickn33/antigravity-awesome-skills
Web App (Skill Browser)	https://sickn33.github.io/antigravity-awesome-skills/
Getting Started Guide	https://github.com/sickn33/antigravity-awesome-skills/blob/main/docs/users/getting-started.md
Full Usage Guide	https://github.com/sickn33/antigravity-awesome-skills/blob/main/docs/users/usage.md
Role-Based Bundles	https://github.com/sickn33/antigravity-awesome-skills/blob/main/docs/users/bundles.md
Skill Catalog	https://github.com/sickn33/antigravity-awesome-skills/blob/main/CATALOG.md
Contributing Guide	https://github.com/sickn33/antigravity-awesome-skills/blob/main/CONTRIBUTING.md
Community Discussions	https://github.com/sickn33/antigravity-awesome-skills/discussions
Issue Tracker	https://github.com/sickn33/antigravity-awesome-skills/issues
Changelog	https://github.com/sickn33/antigravity-awesome-skills/blob/main/CHANGELOG.md

12. Summary

Antigravity Awesome Skills is the most comprehensive, cross-tool-compatible collection of agentic skills available for AI coding assistants today — combining 1,239+ battle-tested skills across 9 domains, official contributions from Anthropic, Google, Vercel, and Microsoft, with a one-command install experience and an interactive web browser.

It is best suited for software developers, DevOps engineers, security researchers, and AI/ML practitioners who want to unlock domain-specific, repeatable capabilities in their AI assistant without building skill libraries from scratch.