neo4j-graphrag-skill

Generates, optimizes, and validates Cypher 25 queries for Neo4j 2025.x and 2026.x, covering graph pattern matching, vector and fulltext search, subqueries, batch writes, indexes, and query optimization.

Design, review, and refactor Neo4j graph data models — covers choosing node labels vs relationship types vs properties, migrating relational/document schemas to graph, detecting anti-patterns like supernodes, and enforcing schema with constraints and indexes.

Ingests unstructured and semi-structured documents (PDFs, HTML, plain text, Markdown) into Neo4j as a knowledge graph — covers chunking, LLM entity/relationship extraction, Document-Chunk-Entity graph structures, and RAG pipeline integration.

Creates and manages Neo4j vector indexes for similarity search, supporting HNSW configuration, quantization, embedding storage on nodes and relationships, and both the SEARCH clause and legacy query procedures.

A skill that guides users or agents through 8 sequential stages to go from zero to a running Neo4j application, covering prerequisites, provisioning, data modeling, loading, exploration, querying, and app building.

A Claude Code skill for the Neo4j Python Driver v6, covering driver lifecycle, execute_query, managed and explicit transactions, async patterns, result handling, UNWIND batching, and connection pool tuning for Python applications connecting to Neo4j.

Authoritative reference for the neo4j-agent-memory Python package, a graph-native memory system for AI agents built on Neo4j providing three memory layers (short-term, long-term, reasoning) in a single knowledge graph. Supports integrations with LangChain, PydanticAI, CrewAI, AWS Strands, Google ADK, OpenAI Agents, LlamaIndex, and the hosted NAMS service.

Guides you through Neo4j Graph Data Science (GDS) workflows including graph projection, algorithm execution (PageRank, Louvain, WCC, FastRP, KNN), execution modes (stream/stats/mutate/write), memory estimation, and the GDS Python client for building recommendation pipelines and graph analytics.

Diagnoses and fixes slow Neo4j Cypher queries by interpreting EXPLAIN/PROFILE execution plans, identifying problematic operators like AllNodesScan and CartesianProduct, and prescribing fixes including indexes, query hints, rewrites, and runtime selection.

A comprehensive Neo4j data import skill covering LOAD CSV, CALL IN TRANSACTIONS, neo4j-admin bulk import, and APOC load procedures. It provides method selection guidance, type coercion, null handling, concurrent transactions, pre-import constraint setup, and post-import validation.