The Ultimate AI Agent Template Collection for Gemini, Claude & Cursor. Supercharge your vibe coding with expert Agents, specialized Skills, and automated Workflows.

MeetKit AI is a "Second Brain" for your AI coding assistant. It provides structured knowledge that transforms your AI from a generic helper into a domain expert.

┌──────────────────────────────────────────────────────────┐
│                     MeetKit AI v1.2.2                    │
├──────────────────────────────────────────────────────────┤
│  🤖 20 Domain Agents    │  Specialists for every task   │
│  🧠 47 Core Skills      │  Deep technical knowledge     │
│  ⚡ 16 Smart Workflows  │  Automated task pipelines     │
└──────────────────────────────────────────────────────────┘

npx @meet644/meetkit-ai init

Or install globally:

npm install -g @meet644/meetkit-ai
npx meetkit-ai init

MeetKit AI is designed for vibe coding — the practice of conversationally building software with AI. Here's how to use each component:

Agents are specialist personas your AI can adopt. Mention them to activate expert mode:

You: "Hey @mobile-developer, design a premium login screen for my Flutter app"

AI: 🤖 Applying @mobile-developer expertise...
    [Generates production-grade Flutter code with animations, theming, responsive layout]

Available Agents:

Skills are deep knowledge modules. Reference them for specific technical guidance:

You: "I need to implement state management. Check the flutter-mastery skill."

AI: [Reads .agent/skills/flutter-mastery/SKILL.md]
    [Provides Riverpod architecture with proper dependency injection]

High-Value Skills:

• flutter-mastery — Production Flutter architecture
• typescript-patterns — Advanced type utilities
• aws-cloud — Serverless, Lambda, DynamoDB
• docker-kubernetes — Container orchestration
• ai-prompt-engineering — System prompts, RAG

Workflows are automated pipelines triggered by slash commands:

You: "/plan"
AI: [Activates project-planner agent]
    [Generates structured implementation plan with phases]

You: "/security-audit"  
AI: [Runs OWASP checks, dependency scanning, code review]
    [Generates prioritized findings report]

You: "/refactor"
AI: [Analyzes code complexity]
    [Executes incremental refactoring with tests]

Power Workflows:

• flutter-ui-ux — Animations, theming, responsive design
• flutter-mastery — Clean architecture, Riverpod, testing
• mobile-design — iOS & Android platform patterns

• fastapi-expert — Async APIs, Pydantic v2
• django-expert — Enterprise web development
• python-data-science — ML pipelines, Pandas

• aws-cloud — Lambda, S3, DynamoDB, CloudFormation
• docker-kubernetes — Containers, Helm, K8s
• deployment-procedures — CI/CD, rollback strategies

The system automatically picks the right expert based on your request:

You: "Design a glassmorphic dashboard for my Next.js app"

AI: 🤖 Identifying intent... AI: ⚡ Applying @frontend-specialist + @ui-ux-pro-max...

After running meetkit-ai init, you'll have:

your-project/
└── .agent/
    ├── ARCHITECTURE.md    # System overview
    ├── agents/            # 20 specialist agents
    ├── skills/            # 47 knowledge modules  
    ├── workflows/         # 16 slash command pipelines
    └── scripts/           # Automation scripts

Keep your kit fresh:

meetkit-ai update

Got a better way to do things? PRs are welcome!

Repository: github.com/Meet6338-X/MeetKitAI