This repository contains two services packaged as Podman containers:

1. LangGraph Service (FastAPI AI agent exposing /chat endpoint)
2. Quart App (handles WhatsApp webhooks and forwards messages to LangGraph)

Both services can be built and run locally using Podman. Follow the steps below to get everything up and running.

• Podman installed on your machine

• A project root containing:

  • langgraph-service/ directory
  • meta_app/ directory
  • .env file with your environment variables

Create a .env file in the project root with the following variables:

# Shared variables
APP_SECRET
ACCESS_TOKEN
APP_ID
APP_SECRET
RECIPIENT_WAID
VERSION=v19.0
PHONE_NUMBER_ID
VERIFY_TOKEN
LANGGRAPH_URL=http://langgraph-service:8000

Important: Do not surround values with quotes (") in your .env file.

project-root/
├── langgraph/
│   ├── __pycache__/
│   ├── graphrag_reduced/
│   ├── secrets/
│   ├── src/
│   ├── yamls/
│   ├── .env
│   ├── config.py
│   ├── Dockerfile
│   ├── main.py
│   ├── models.py
│   ├── README.md
│   └── requirements.txt
├── meta_app/
│   ├── app/
│   │   └── utils/whatsapp_utils.py
│   ├── .env
│   ├── Dockerfile
│   ├── requirements.txt
│   ├── run.py
│   ├── .gitignore
│   ├── README.md
│   └── supabase_setup.txt
└── README.md

cd langgraph
podman build -t langgraph-service .

cd meta_-app
podman build -t meta_app .

Create a user-defined Podman network to allow containers to communicate by name:

podman network create meta-net

podman run -d \
  --name langgraph-service \
  --env-file .env \
  --network meta-net \
  -p 8000:8000 \
  langgraph-service

podman run -d \
  --name meta_app \
  --env-file .env \
  --network meta-net \
  -p 8080:8080 \
  meta_app

Now:

• LangGraph API is at http://localhost:8000/chat
• Quart App (WhatsApp webhook) is at http://localhost:8080 (adjust your webhook route if different)

curl -X POST http://localhost:8000/chat \
  -H "Content-Type: application/json" \
  -d '{"wa_id":"test","message":"Hola"}'