A curated collection of 130+ production-ready Gen AI apps, agents, installable Agent Skills, MCP servers, and latest-model cookbooks. Built with Claude, GPT-5, Gemini 3, Qwen, GLM, the Model Context Protocol, LangChain, RAG, and Multi-Agent Teams.
- 🧩 Installable Agent Skills - Reusable
SKILL.mdcapabilities for Claude Code, Cowork, and Codex — from landing pages and talking avatars to prompt optimization and QA review - 🔌 MCP-Native - Model Context Protocol servers, clients, and integrations you can plug straight into your agents
- 🧠 Latest Models, Day One - Hands-on cookbooks for Claude Opus 4.8, GPT-5.5, Gemini 3, Qwen3.6, GLM-5, Kimi K2.7, DeepSeek V4, and more
- 💡 Learn by Building - 130+ production-ready applications and experiments you can run, modify, and learn from
- 🎓 From Beginner to Advanced - Structured learning path with starter, intermediate, and advanced projects
- 🌍 Multi-Language Support - Projects supporting English, Hindi, and other Indian languages
- 📚 100+ Libraries Covered - Tutorials across LangChain, LlamaIndex, CrewAI, AG2, Weaviate, and 90+ more AI/ML libraries
Installable, model-agnostic Agent Skills — each a self-contained SKILL.md (plus optional scripts/, references/, and assets/) that teaches an agent a repeatable workflow. Drop them into Claude Code, Cowork, or Codex. See the Skills index → for full details.
- Landing Page Generator - High-converting landing pages as production HTML with PAS/AIDA/BAB copy frameworks, CTA strategy, SEO meta, and conversion/speed audits
- Talking Avatar - Realtime voice-chat apps with a lip-synced character avatar on OpenAI Realtime (Vite/Next.js, BYOK)
- Crazy Ecommerce Builder - Turn a brand brief into an art-directed storefront with original ImageGen product photography
🖥️ Frontend Skill Pack — skills/frontend-skills/
- Frontend Core & Tailwind Component Factory
- Style systems: Glass UI, Neo-Brutalism, Minimal Luxury, Bold SaaS, Editorial, Retro-Futurist
🛠️ Backend Skill Pack — skills/backend-skills/
- MCP Server Builder - Scaffold production MCP servers
- Next.js Route Handler, MERN Auth Best Practices, Mongoose Schema Architect
📝 Docs & Research Skill Pack — skills/docs-writing-research-skills/
The Model Context Protocol is the open standard for connecting agents to tools and data. This repo ships MCP servers, clients, integrations, and runnable examples under mcp/.
- Launch MCP 🚀 - Zero-dependency MCP server that turns any GitHub repo into a complete launch kit: repo analysis, release notes, changelog, blog post, X/Reddit/HN/LinkedIn posts, an interactive HTML demo, and an SVG share card. Tools:
analyze_repo,generate_share_card. - MCP-Use App - Reference client showing how to wire MCP tools into an agent loop
- MCP Workshop - Guided, hands-on notebook for building your first MCP server and client
- AG2 + MCP Client - Multi-agent GroupChat with tool use over MCP
Contributing an MCP project? Servers go in
mcp/servers/, clients inmcp/clients/, integrations inmcp/integrations/, and runnable demos inmcp/examples/.
Day-one testing notebooks for the newest frontier and open models, organized by provider under cookbooks/models/.
- Gemini 3.5 Flash · 3.1 Pro Preview · 3.1 Flash Live · 3.1 Flash Lite
- Gemini 3.0 Structured Output · Embedding 2 · Code Execution
- Moonshot Kimi: K2.7 Code · K2.6
- DeepSeek: V4 Pro · V4 Flash
- Grok Code Fast · MiniMax M2 · Mistral Devstral · Voxtral
- Indic & Multilingual: Sarvam 30B/105B · Sutra · Llama 3 on Indic
Function-calling, tool orchestration, and model-side tool use — the plumbing that turns a chat model into an agent.
- Tool Use Validator - Validate function-calling JSON payloads against a schema before execution
- Prompt Optimizer (CoT) - Rewrite raw tasks into robust Chain-of-Thought prompts with verification
- Agent Output Critic - Strict QA review for hallucinations, security, and logic flaws before delivery
- Git Conventional Commits - Generate Conventional Commits and PR writeups from real diffs
- Function Calling with Gemma - Local tool-calling with open Gemma models
- Gadget Comparator (Gemini URL Context) - Grounded product analysis using Gemini's URL-context tool
Perfect for beginners getting started with Gen AI:
- LangChain Basics - Build intelligent workflows with LangChain
- Fine-Tuning with Nebius Token Factory - LoRA fine-tuning for custom LLMs
- Getting Started with Pydantic AI - Type-safe AI development
- CrewAI Essentials - Quick start guide for multi-agent systems
- Hugging Face Transformers - Foundation for Gen AI and NLP
- LlamaIndex - Intelligent data integration for LLMs
- ChromaDB - Vector database for embeddings
Build more complex AI systems:
- AutoGen Multi-Agent System - Collaborative AI agents
- AG2 Multi-Agent AI Systems - GroupChat, tool use, MCP client with AG2 (formerly AutoGen)
- LangGraph Multi-Agent Swarm - Advanced agent orchestration
- CSV Agents - Data analysis with LangChain & LlamaIndex
- AI Customer Support Agent - Production-ready support system
- RAG Systems - Advanced retrieval-augmented generation
Production-grade implementations and cutting-edge research:
- Cerebras Inference Comparison - High-performance model benchmarking
- Deep Stock Research Agent - Multi-step autonomous financial research
- Model Evaluation Cookbooks - Comprehensive provider-by-provider model testing
- Advanced RAG Architectures - Sophisticated retrieval systems
- Chat with QWEN3 Coder - Advanced coding assistant powered by QWEN
- Chat with PDF or Webpage - Extract and interact with document content
- Chat with GPT-OSS - Open-source GPT interface
- Sutra V2 Multilingual Chatbot - Support for Hindi and Indian languages
- Vibe Voice TTS - Text-to-speech application
- Chess Playing Agents - AI chess opponents with GLM model
- OpenAI Gemini Chess - Multi-model chess game
- QWEN Game Generator - Automated game creation
- World's Fastest Game Generator - Rapid game development with QWEN3
- Multilingual Quiz Generator - Create quizzes in multiple languages
- QnA Generator - Automated question generation
- Origami Tutorial Generator - Creative step-by-step tutorials
- Indian Language Quiz - Regional language support
- Language Learner - Interactive language learning platform
- Visual Question Generator - Image-based question creation
- AI Business Consultant - Strategic business advisor
- AI Ad Generator - Marketing content creation
- Finance AgentOS - Financial analysis and insights
- Stock Market Agent - Real-time market analysis
- Deep Stock Research Agent - Advanced financial research
- Marketing Automation - Automated marketing campaigns
- Perplexity AI Research Assistant - AI-powered research automation
- Cerebras Search - Ultra-fast search engine
- Personalized Search Agent - Customized search experience
- Similarity Analyzer - Content comparison with Venn diagrams
- GitHub README Generator - Auto-generate documentation
- Browser Automation - Web automation interface
- World's Fastest Website Generator - Instant web development
- Book Writer AI - Automated content creation
- LLM-friendly Web Scraper - Intelligent web scraping
- Gemini 2.0 Multimodal - Google's multimodal AI
- News to Blog Automator - Content transformation pipeline
- Gadget Comparator - Product analysis with Gemini
- Chroma Cloud RAG - Cloud-based retrieval system
- MCP Implementation - Model Context Protocol
- TypeSafe Agno - Type-safe AI development
- NextJS Image Workflow - Advanced image processing
Quick directory map to help contributors and learners navigate faster.
- Definition: Installable Agent Skills — each a
SKILL.md(plus optionalscripts/,references/,assets/) that teaches an agent a repeatable workflow. - Top Skills:
- Definition: Model Context Protocol servers, clients, integrations, and runnable examples.
- Top Resources:
- Definition: Autonomous agent projects that choose their own tools and actions. (Open for contributions.)
- Definition: Predefined, multi-step AI automation pipelines. (Open for contributions.)
- Definition: Runnable, user-facing GenAI applications and demos (~60 projects).
- Top Apps:
- Definition: Educational notebooks — organized into
models/,tools/,agents/,mcp/,rag/,multimodal/,fine-tuning/,evaluations/, andworkshops/. - Top Cookbooks:
- Definition: Agent-building notebooks (single-agent and multi-agent) with practical orchestration patterns.
- Top Cookbooks:
- Definition: Retrieval-Augmented Generation implementations, multimodal retrieval, and document-grounded QA.
- Top Cookbooks:
- Definition: Hands-on workshop notebooks and guided classroom-style build sessions.
- Top Cookbooks:
- Definition: Older, superseded, or reference content kept for posterity — including the original
100-os-libraries/(100+ library cookbooks),experiments/, androadmaps/. - Top Resources:
Curated from recently updated and high-interest notebooks in this repository.
- Claude Opus 4.8 Cookbook
- GPT-5.5 Cookbook
- Gemini 3.5 Flash Cookbook
- Qwen3.6 Max Preview Cookbook
- GLM OCR Cookbook
- Kimi K2.7 Code Cookbook
- DeepSeek V4 Pro Cookbook
- How to Build Claude-Powered RAG from Scratch
- Kimi K2.5 Agent Swarm Cookbook
- MCP Workshop
- Python 3.8+
- pip or conda package manager
- Node.js 18+ (for MCP servers and JS/TS apps)
- API keys for services you want to use (OpenAI, Anthropic, Google, etc.)
-
Clone the repository
git clone https://github.com/buildfastwithai/gen-ai-experiments.git cd gen-ai-experiments -
Navigate to your desired project
cd apps/your-desired-project -
Install dependencies
pip install -r requirements.txt
-
Set up API keys
# Create a .env file and add your API keys echo "OPENAI_API_KEY=your_key_here" > .env
-
Run the application
streamlit run app.py # or python app.py # or jupyter notebook
-
Follow project-specific instructions in each project's
README.mdfor detailed setup
Agent Skills are model-agnostic. Point your agent (Claude Code, Cowork, or Codex) at any folder under skills/ and it will read the SKILL.md and follow the workflow. See the Skills index for what each one does.
# Example: Launch MCP in Claude Code
/plugin marketplace add buildfastwithai/launch-mcp
/plugin install launch-mcp@buildfastwithai-pluginsWe welcome contributions from the community! Here's how you can help:
- Fork the repository
- Create a feature branch (
git checkout -b feature/AmazingFeature) - Commit your changes (
git commit -m 'Add some AmazingFeature') - Push to the branch (
git push origin feature/AmazingFeature) - Open a Pull Request
- Skills →
skills/<skill-name>/with aSKILL.md - MCP →
mcp/servers/,mcp/clients/,mcp/integrations/, ormcp/examples/ - Apps →
apps/<app-name>/with aREADME.mdandrequirements.txt - Agents →
agents/<agent-name>/ - Workflows →
workflows/<workflow-name>/ - Cookbooks →
cookbooks/<category>/ - Legacy / experimental →
archive/legacy/
- Follow the existing project structure
- Include a detailed
README.mdfor new projects - Add requirements.txt with all dependencies
- Test your code before submitting
- Write clear commit messages
- Update documentation as needed
Found a bug or have a feature request? Please create a GitHub Issue with:
- Clear description of the issue/feature
- Steps to reproduce (for bugs)
- Expected vs actual behavior
- Screenshots (if applicable)
Show your support by starring this repository!
Made with ❤️ by BuildFastWithAI
⭐ Star this repo if you find it helpful!