Agent Skills & Coding Tools

AI agents and coding tools become significantly more effective when they have direct access to your Neo4j graph database. Rather than relying solely on training data, agents can retrieve live schema information, execute and validate Cypher queries, and use real graph data to generate code, tests, and integrations — all without leaving the editor.

Neo4j supports two complementary integration layers:

  • Neo4j Agent Skills — Installable knowledge bundles that teach AI agents Neo4j-specific patterns. 20+ skills covering Cypher, GraphRAG, drivers, GDS, data import, agent memory, and more. Works with Claude Code, Cursor, Cline, Gemini CLI, Codex, and any Agent Skills-compliant tool.

  • Coding Tool Integrations — The Neo4j MCP server and dedicated editor extensions connect agents directly to a live Neo4j database for schema inspection, Cypher execution, and real-data validation. Supported in VS Code, Gemini CLI, Antigravity, Kiro, Claude Code, Cursor, and more.

Neo4j Agent Skills

Browse and install at skills.sh/neo4j-contrib/neo4j-skills.

neo4j-getting-started-skill

The flagship skill: takes you from zero to a running graph application in a single Claude Code session. Invoke it with a plain-language description — it provisions a database, designs the schema, loads data, and generates a runnable app end-to-end.

/neo4j-getting-started-skill fraud detection for a fintech startup
/neo4j-getting-started-skill healthcare patient graph, local Docker, FastAPI, synthetic data

Install: npx skills add neo4j-contrib/neo4j-skills/neo4j-getting-started-skill

Quick Install

Agent Install command

Claude Code

/plugin marketplace add https://github.com/neo4j-contrib/neo4j-skills.git

Gemini CLI

gemini extensions install https://github.com/neo4j-contrib/neo4j-skills

Cursor / Cline / Windsurf

npx skills add neo4j-contrib/neo4j-skills

Codex

Clone repo, copy to ~/.codex/plugins/ — see install guide

Coding Tool Integrations

The Neo4j MCP server gives any MCP-compatible coding tool live access to your graph:

  • Schema-aware code generation — retrieve node labels, relationship types, and property keys to generate Pydantic models, GraphQL schemas, Java POJOs, or TypeScript interfaces matching your actual graph

  • Cypher validation — run generated queries against the real database before shipping; catch syntax errors, full scans, and logic bugs immediately

  • Data-driven test generation — sample real records to produce realistic integration tests and fixtures

Supported editors: VS Code, Claude Code, Claude Desktop, Cursor, Windsurf, Kiro, Antigravity, Cline, Gemini CLI, and any MCP-compatible client.

See Coding Tool Integrations for editor-specific configuration.

GraphAcademy Courses

If you want to learn how LLMs and Knowledge Graphs combine to improve GenAI applications, check out the Neo4j & LLM courses on GraphAcademy.

llm fundamentals