Redis + LangGraph Example

A full-stack production example with Redis broker/event bus, PostgreSQL storage, and a LangGraph research pipeline agent.

Quick Start

cd examples/redis_langgraph
docker compose up --build

Then open http://localhost:8000/chat in your browser.

What It Demonstrates

1. Redis as broker, event bus, AND cancel registry — horizontally scalable task processing
2. PostgreSQL as durable storage — tasks survive server restarts
3. LangGraph StateGraph with 3 nodes (research → analyze → summarize) and custom stream events
4. Cooperative cancellation inside graph nodes via ctx.is_cancelled
5. Streaming artifact chunks from LangGraph → A2A SSE
6. Built-in debug UI at /chat

Architecture

┌──────────────┐     ┌──────────────┐     ┌──────────────┐
│  Agent (N)   │────▶│    Redis     │◀────│  Agent (N)   │
│  uvicorn     │     │  Streams +   │     │  uvicorn     │
│              │     │  Pub/Sub     │     │              │
└──────┬───────┘     └──────────────┘     └──────┬───────┘
       │                                         │
       └────────────┬────────────────────────────┘
                    ▼
           ┌──────────────┐
           │  PostgreSQL  │
           │  (storage)   │
           └──────────────┘

Server Code

from a2akit import A2AServer, AgentCardConfig, CapabilitiesConfig

server = A2AServer(
    worker=ResearchWorker(),
    agent_card=AgentCardConfig(
        name="Research Pipeline Agent",
        description="Multi-stage research agent powered by LangGraph.",
        version="1.0.0",
        capabilities=CapabilitiesConfig(streaming=True),
    ),
    broker="redis://redis:6379/0",          # RedisBroker
    event_bus="redis://redis:6379/0",       # RedisEventBus
    # cancel_registry auto-created as RedisCancelRegistry
    storage="postgresql+asyncpg://...",     # PostgreSQLStorage
)

Scaling

# Run 3 worker replicas sharing the same Redis broker
docker compose up --build --scale agent=3

Each replica joins the same Redis consumer group — tasks are load-balanced automatically.

CLI Client

python client.py "Research the current state of AI agent frameworks"

Files

File	Description
server.py	A2AServer with LangGraph worker
client.py	CLI client for terminal usage
Dockerfile	Python image with dependencies
docker-compose.yml	Redis + Postgres + Agent
requirements.txt	pip dependencies