Create Context Graph
Create Context Graph is an interactive CLI scaffolding tool that generates complete, domain-specific context graph applications. Think of it as create-next-app, but for AI agents backed by graph memory.
Given a domain (like healthcare, financial services, or wildlife management) and an agent framework, it generates a full-stack application: a FastAPI backend with a configured AI agent, a Next.js + Chakra UI frontend with NVL graph visualization, a Neo4j schema with synthetic data, and domain-specific tools that let the agent query and reason over your knowledge graph.
Key Features
- 22 built-in domains -- healthcare, financial services, real estate, manufacturing, scientific research, software engineering, and more. Each domain ships with a complete ontology, agent tools, demo scenarios, and fixture data.
- 8 agent frameworks -- PydanticAI, Claude Agent SDK, OpenAI Agents SDK, LangGraph, CrewAI, Strands, Google ADK, and Anthropic Tools. Pick the one you know, or try something new.
- Multi-turn conversations -- every generated agent uses neo4j-agent-memory for conversation persistence. Session history is stored in Neo4j and retrieved on each turn, so follow-up questions work naturally.
- Graph-native AI agents -- every generated agent comes with Cypher-powered tools for querying entities, relationships, and decision traces in Neo4j. Tool calls stream in real-time with live progress indicators.
- Streaming chat -- responses stream token-by-token via Server-Sent Events. Tool calls appear as a live timeline with spinner indicators as each executes. The graph visualization updates incrementally after each tool completes, not just at the end.
- Interactive graph visualization -- the frontend includes an NVL-powered graph explorer with entity detail panel (click any node to see all properties and connections), a document browser with template filtering, and a decision trace viewer.
- Rich demo data -- each domain ships with LLM-generated fixture data: 80-90 entities with realistic names, 25+ professional documents (discharge summaries, trade confirmations, lab reports), and 3-5 multi-step decision traces. All loaded into Neo4j via
make seedand browsable in the frontend. - Flexible Neo4j setup -- connect to Neo4j Aura (free cloud tier with
.envimport), run locally with@johnymontana/neo4j-local(no Docker needed), use Docker Compose, or connect to any existing instance. - SaaS data import -- connect Gmail, Slack, Jira, GitHub, Notion, and Salesforce to populate your graph with real data.
- Custom domains -- describe your domain and let the tool generate a complete ontology, or write your own YAML definition from scratch.
Quick Install
No installation required. Run directly with uvx (Python) or npx (Node.js):
# Python (recommended)
uvx create-context-graph
# Node.js
npx create-context-graph
Quick Start (Non-Interactive)
Skip the wizard entirely by passing flags:
uvx create-context-graph my-app \
--domain healthcare \
--framework pydanticai \
--demo-data
This creates a my-app/ directory with a complete healthcare context graph application using PydanticAI as the agent framework, pre-loaded with demo data.
See All Available Domains
uvx create-context-graph --list-domains
What's Next
- Your First Context Graph App -- step-by-step tutorial to create, run, and explore a generated application.
- Customizing Your Domain Ontology -- learn how to modify entity types, relationships, and agent tools in your domain YAML.