java-pro

This repository contains 300+ specialized skills across software development, operations, security, and business domains. Each skill is a directory-based package that teaches Antigravity's agent how to perform specific tasks.

What's Included?

The skills are derived from three types of Claude Code components, all unified into the Antigravity Skill format:

1. Domain Skills (e.g., k8s-manifest-generator, async-python-patterns): Specialized knowledge packages.
2. Specialist Agents (e.g., backend-architect, security-auditor): Persona-based instruction sets for complex reasoning.
3. Commands & Workflows (e.g., full-stack-orchestration-full-stack-feature, conductor-implement): Structured, multi-step procedures.

---

Skills are flattened in the skills/ directory, but cover these broad categories:

💻 Development & Languages

• Python: python-pro, fastapi-pro, async-python-patterns, uv-package-manager
• JavaScript/TypeScript: typescript-pro, react-modernization, nextjs-app-router-patterns
• Systems: rust-pro, golang-pro, memory-safety-patterns
• Mobile: frontend-mobile-development-component-scaffold, react-native-architecture

☁️ Infrastructure & Operations

• Kubernetes: kubernetes-architect, helm-chart-scaffolding, gitops-workflow
• Cloud: cloud-architect, terraform-module-library, cost-optimization
• CI/CD: cicd-automation-workflow-automate, github-actions-templates, gitlab-ci-patterns

🔒 Security & Quality

• Security: security-auditor, sast-configuration, security-scanning-security-hardening
• Code Quality: code-review-ai-ai-review, code-refactoring-refactor-clean, codebase-cleanup-tech-debt
• Testing: unit-testing-test-generate, tdd-workflows-tdd-cycle, e2e-testing-patterns

🔄 Workflows & Architecture

• Conductor: conductor-implement, context-driven-development (Context-Driven Development)
• Architecture: c4-architecture-c4-architecture, microservices-patterns, api-design-principles
• Orchestration: full-stack-orchestration-full-stack-feature, incident-response-incident-response

📊 Data & AI

• Data Engineering: data-engineer, spark-optimization, dbt-transformation-patterns
• AI/ML: ml-pipeline-workflow, prompt-engineering-patterns, rag-implementation

---

Antigravity loads metadata (name + description) from every installed skill at session start. More skills means more token usage and a higher chance of irrelevant auto-activation. Prefer targeted installs via search, tags, or bundles. install --all is advanced and not recommended for most projects.

This repo ships a generated catalog for discovery:

• CATALOG.md (human-readable index)
• catalog.json (machine-readable index used by the CLI)
• bundles.json (curated bundles)
• aliases.json (short names that map to long skill IDs)

Regenerate after adding/editing skills:

```bash

npm run build:catalog

```

When a conversation starts, Antigravity loads the metadata (name & description) from all skills.

ANTIGRAVITY automatically activates a skill when your request matches its description.

Examples:

• "Help me design a REST API for a user service" → Activates api-design-principles and backend-architect.
• "Scaffold a new FastAPI project" → Activates python-development-python-scaffold.
• "Review this PR for security issues" → Activates `security-scanning-security-