Jest
LangGraph is a framework for building stateful, multi-step AI agent workflows as directed graphs, with a Pregel-based runtime engine for orchestrating LLM calls, tool usage, and human-in-the-loop interactions. It includes a core graph execution engine, persistent checkpointing backends (PostgreSQL, SQLite, in-memory), prebuilt agent patterns (ReAct agents, tool nodes), a Python SDK client for a remote LangGraph API server, and a CLI for building/deploying LangGraph applications as Docker containers.
Tech Stack
JestReactRedisPostgreSQL
Key Features
- StateGraph & Pregel Runtime
- Pluggable Checkpoint Persistence
- Prebuilt ReAct Agent
- Human-in-the-Loop Interrupts
- BaseStore with Vector Search
- Python SDK Client
- CLI for Docker Deployment
- Managed Values (SharedValue, IsLastStep)
Coding Conventions
Standards and patterns used in this codebase
configuration
Configuration is centralized in config files
Config files: libs/cli/langgraph_cli/config.py, libs/langgraph/langgraph/config.pylibs/cli/langgraph_cli/config.pylibs/langgraph/langgraph/config.py
data-layer
Database operations are in data layer
Data files: libs/cli/generate_schema.py, libs/cli/langgraph_cli/schemas.pylibs/cli/generate_schema.pylibs/cli/langgraph_cli/schemas.py
file-organization
Standard project structure
Code is organized by feature/moduleexamples/chatbot-simulation-evaluation/simulation_utils.pylibs/cli/generate_schema.pylibs/checkpoint/tests/test_memory.py