Personal collection of explorations, experiments, and prototypes in the Agentic Workflow domain.
This project serves as a learning platform for understanding and building intelligent agents using LangChain and related libraries. It combines both example explorations of library usage and practical agent implementations.
agent-lab/
├── langchain-examples/ # Examples exploring LangChain library usage
├── langgraph-examples/ # Examples exploring LangGraph library usage
├── deepagent-examples/ # Examples exploring DeepAgent library usage
├── projects/ # Practical agent demo projects
│ ├── deepsearch-demo/ # Deep search agent demo
│ └── dataagent-demo/ # Data agent demo
├── pyproject.toml # Workspace configuration
└── README.md # This file
The *-examples directories contain exploratory code for understanding how to use various libraries in the LangChain ecosystem:
langchain-examples/- Core LangChain functionality (models, prompts, chains, agents, middleware)langgraph-examples/- State graph-based agent workflowsdeepagent-examples/- DeepAgent framework usage
The projects/ directory contains complete, runnable agent demo implementations that showcase real-world patterns and use cases.
This project uses uv for dependency management and virtual environment handling.
- Python 3.12+
- uv
- Clone the repository
- Sync the workspace:
uv syncDemo projects are defined as optional dependencies. Install them using extras:
# Install specific demo projects
uv sync --extra deepsearch
uv sync --extra dataagent
# Install multiple demo projects
uv sync --extra deepsearch --extra dataagent
# Install all demo projects
uv sync --all-extrasOnce a demo project is installed via extras, you can run it using:
uv run deepsearch-demo
uv run dataagent-demoExample scripts can be run directly:
uv run langchain-examples/model_stream_output.py