langX101

A comprehensive repository for testing and exploring GenAI Observability, AI Agent Frameworks, Model Context Protocol (MCP), and Agent-to-Agent (A2A) Protocol implementations.

🎯 Overview

This repository serves as a hands-on laboratory for experimenting with various generative AI tools, observability platforms, and integration patterns. It contains practical examples, tutorials, and implementations across multiple AI frameworks and monitoring solutions.

📁 Repository Structure

🤖 AI Agents & Multi-Agent Systems

Directory	Description
a2a/	Agent-to-Agent (A2A) protocol implementations
a2a_langgraph_mcp/	LangGraph with MCP integration for multi-agent systems
adk/	Agent Development Kit with MCP server implementations
swarm/	OpenAI Swarm agent implementations
oai-agent/	OpenAI agent examples with function calling
crew/	CrewAI multi-agent orchestration examples
langgraph/	LangGraph stateful agent workflows
agent-semantic-convention/	OpenTelemetry semantic conventions for AI agents

🔧 Model Context Protocol (MCP)

Directory	Description
mcp/	Core MCP implementations (weather, email, tutorial servers)
mcp-client/	MCP client implementations
mcp-go/	Go-based MCP server and client examples
oai-mcp/	OpenAI integration with MCP (filesystem, SSE)
langchain-mcp/	LangChain integration with MCP servers
fastmcp/	FastMCP server implementations

🦜 LangChain Ecosystem

Directory	Description
langchain/	Comprehensive LangChain examples (callbacks, RAG, AutoGPT)
langchainjs/	LangChain JavaScript/TypeScript examples
langgraph/	LangGraph stateful workflows and agents
langserve/	LangServe server implementations
langflow/	LangFlow visual workflows and custom components
langsmith/	LangSmith evaluation and monitoring
langfuse/	LangFuse observability integration

📊 Observability & Monitoring

Directory	Description
otel/	OpenTelemetry instrumentation for AI frameworks
traceloop/	Traceloop SDK integration examples
arize/	Arize AI monitoring integration
helicone/	Helicone observability examples
langtrace/	LangTrace monitoring implementation
llmonitor/	LLM monitoring examples
newrelic/	New Relic AI monitoring
promptlayer/	PromptLayer integration

OpenTelemetry (otel/) Includes:

• OpenAI auto-instrumentation
• LangChain instrumentation
• ChromaDB instrumentation
• WatsonX instrumentation
• Custom span and trace examples

☁️ Cloud Providers & Models

Directory	Description
aws/	AWS Bedrock examples and model implementations
watsonx/	IBM WatsonX examples with RAG
openai/	OpenAI API examples and assistants
deepseek/	DeepSeek model implementations
litellm/	LiteLLM proxy examples
llamastack/	Meta Llama Stack implementations

🗄️ Vector Databases & Memory

Directory	Description
milvus/	Milvus vector database examples
embedchain/	Embedchain implementations
mem0/	Mem0 memory layer for AI agents
postgres/	PostgreSQL with pgvector examples

🛡️ Security & Evaluation

Directory	Description
llmguard/	LLM security and guardrails
eval/	Model evaluation frameworks

🌐 Web & API Development

Directory	Description
graphql_instana/	GraphQL with Instana monitoring
graphql-example/	GraphQL examples
my_flask_graphql_app/	Flask GraphQL application
streamlit-test/	Streamlit application examples
drizzle-demo/	Drizzle ORM demo

🔧 Utilities & Tools

Directory	Description
python/	Python utilities and decorators
mylib/	Custom library implementations
autowrapt/	Auto-wrapping utilities
oauth/	OAuth authentication examples
cline/	Cline AI coding assistant examples
scira/	Scira search integration

🚀 Getting Started

Prerequisites

• Python 3.8+
• Node.js (for TypeScript/JavaScript examples)
• Go (for Go examples)
• Docker (for some services)
• uv (recommended for Python dependency management)

Installation

1. Clone the repository:

git clone https://github.com/gyliu513/langX101.git
cd langX101

2. Set up Python environment (using uv):

uv venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate
uv sync

Or using pip:

python -m venv venv
source venv/bin/activate
pip install -r requirements.txt  # if available in specific directory

3. Configure environment variables:

cp dot-env.txt .env
# Edit .env with your API keys

📚 Key Examples

MCP Server Setup

Check out the MCP examples in /mcp/ for setting up Model Context Protocol servers:

• Weather server implementation
• Email sending capabilities
• Multi-tool scenarios

Agent-to-Agent (A2A) Communication

Explore the /a2a/ and /a2a_langgraph_mcp/ directories for:

• Agent orchestration patterns
• Multi-agent coordination
• MCP + A2A integration

OpenTelemetry GenAI Observability

The /otel/ directory contains comprehensive examples for:

• Auto-instrumentation of LLM calls
• Custom span attributes for AI operations
• Integration with various observability backends

RAG Implementations

Find various RAG patterns:

• WatsonX RAG in /watsonx/
• LangChain RAG examples in /langchain/
• Llama Stack RAG in /llamastack/
• Vector database integrations

🔗 Related Resources

• Model Context Protocol (MCP) Documentation
• OpenTelemetry GenAI Semantic Conventions
• LangChain Documentation
• LangGraph Documentation
• Traceloop OpenLLMetry

👤 Author

Guangya Liu - Senior Technical Staff Member at IBM

• OpenTelemetry GenAI Semantic Convention Maintainer
• Apache Mesos PMC Member & Committer
• IBM Master Inventor with 30+ patents
• LinkedIn | GitHub | Medium

🤝 Contributing

Contributions are welcome! Feel free to:

• Add new tool integrations
• Improve existing examples
• Share interesting use cases
• Report issues or suggestions

📄 License

This project is licensed under the Apache 2.0 License - see the LICENSE file for details.

🏷️ Tags

genai observability mcp a2a langchain langgraph openai agents rag monitoring opentelemetry ai-agents multi-agent traceloop