dev kit with neo4j (#2)

Co-authored-by: pavanmantha <[email protected]>

Author: pavanjava

.env

@@ -3,4 +3,6 @@ POSTGRES_PASSWORD=password
 POSTGRES_DB=n8n
 
 N8N_ENCRYPTION_KEY=s3cr3tk3y
-N8N_USER_MANAGEMENT_JWT_SECRET=jwts3cr3t
+N8N_USER_MANAGEMENT_JWT_SECRET=jwts3cr3t
+
+NEO4J_PASSWORD=p@55w0rd

README.md

@@ -5,21 +5,23 @@ This is a self hosting repository for creating AI Agents and AI Agent powered wo
 ![Arch](architecture.png)
 
 ### How to run
-- `git clone <repo>`
+- `git clone [email protected]:pavanjava/ai-agents-dev-kit.git`
 - `cd <repo>`
 - `sh setup.sh`
 
 ### What all gets installed as part of devkit?
 
 - ✨&nbsp; n8n platform
 - ✨&nbsp; Qdrant version 1.13 (latest)
+- ✨&nbsp; Neo4j for knowledge graphs
 - ✨&nbsp; Ollama with llama3.2: latest installed
 - ✨&nbsp; Redis stack for fully function redis cache
 - ✨&nbsp; PostgreSQL for maintaining n8n state, data and configuration
 
 ### What all you can use this platform for ?
 - 🌟&nbsp; Intelligent AI Agents
 - 🌟&nbsp; Intelligent Workflow management
+- 🌟&nbsp; Intelligent knowledge graph Agents
 - 🌟&nbsp; Intelligent RAG pipeline Agents
 - 🌟&nbsp; Intelligent social media Agents
 - 🌟&nbsp; much more

ai-agents-dev-kit.yml

@@ -4,6 +4,7 @@ volumes:
   agent_llm_data: 
   agent_vector_data:
   agent_cache_data:
+  agent_neo4j_data:
 
 networks:
   aiagent_net:
@@ -96,4 +97,25 @@ services:
       - "-c"
       - "sleep 3; OLLAMA_HOST=agent-llm:11434 ollama pull llama3.2"
     depends_on:
-      - ollama
+      - ollama
+
+  neo4j:
+    image: neo4j:5.26-enterprise
+    container_name: agent-graph-db
+    networks: [ 'aiagent_net' ]
+    restart: unless-stopped
+    ports:
+      - 7474:7474  # HTTP
+      - 7687:7687  # Bolt
+    environment:
+      - NEO4J_AUTH=neo4j/${NEO4J_PASSWORD}
+      - NEO4J_ACCEPT_LICENSE_AGREEMENT=yes
+      - NEO4J_server_memory_heap_initial__size=1G
+      - NEO4J_server_memory_heap_max__size=2G
+    volumes:
+      - agent_neo4j_data:/data
+    healthcheck:
+      test: [ "CMD", "wget", "--no-verbose", "--tries=1", "--spider", "http://localhost:7474" ]
+      interval: 10s
+      timeout: 5s
+      retries: 5

architecture.png

@@ -6,7 +6,7 @@ docker compose -f ai-agents-dev-kit.yml up -d --scale ollama-init=0
 
 # Step 2: Run ollama-init as a one-off task and wait for completion
 echo "Running ollama-init task..."
-docker compose run --rm ollama-init
+docker compose -f ai-agents-dev-kit.yml run --rm ollama-init
 
 # Step 3: Log completion
 echo "ollama-init task completed and container removed."