skills.rest
cardox6
cardox6COMMUNITY
Explore all skills by this author →
GitHub

neo4j-document-import-skill

Ingest unstructured documents into Neo4j as knowledge graphs using LLM extraction.

Updated Aug 25, 2026
Search Tags:#knowledge graph#llm#chunking#document ingestion#neo4j#entity extraction#graphrag
One-click install
npx skills add https://github.com/cardox6/steuer-graph --skill neo4j-document-import-skill-cardox6
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: neo4j-document-import-skill
Source: https://github.com/cardox6/steuer-graph/tree/main/.agents/skills/neo4j-document-import-skill
Command: npx skills add https://github.com/cardox6/steuer-graph --skill neo4j-document-import-skill-cardox6

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires neo4j-graphrag, neo4j, and includes references (resource) components.

What problem does it solve? Turning PDFs, Markdown, HTML, and JSON into a queryable Neo4j knowledge graph requires coordinating chunking, LLM entity extraction, embeddings, and entity resolution, which is error-prone to wire up from scratch. ## Core Features & Use Cases - SimpleKGPipeline orchestration: Configure neo4j-graphrag's SimpleKGPipeline with schema definitions, chunking strategies, and multiple LLM providers (OpenAI, Anthropic, Vertex AI, Ollama, Bedrock). - Entity resolution and schema design: Merge duplicate entities with exact, fuzzy, or semantic resolvers, and define GraphSchema with constraints for production-quality extraction. - Alternative ingestion paths: Use the no-code LLM Graph Builder web UI, apoc.load.json for semi-structured JSON, or LangChain/LlamaIndex document loaders. - Use Case: Ingest a folder of quarterly PDF reports into Neo4j, producing Document→Chunk→Entity graph structures with embeddings ready for GraphRAG retrieval. ## Quick Start Ask the agent to ingest a PDF file into Neo4j using SimpleKGPipeline with a defined entity schema and run entity resolution afterward.

Frequently Asked Questions about neo4j-document-import-skill

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I import PDF documents into Neo4j as a knowledge graph?▼

Use SimpleKGPipeline from the neo4j-graphrag Python package with from_file=True and pass the PDF path to run_async. The pipeline chunks the document, extracts entities with an LLM, embeds chunks, and writes Document, Chunk, and entity nodes to Neo4j.

What is the difference between SimpleKGPipeline and LLM Graph Builder?▼

SimpleKGPipeline is a programmatic Python API in neo4j-graphrag for scripted ingestion pipelines. LLM Graph Builder is a no-code web UI for drag-and-drop ingestion, suited to non-developers and rapid prototyping.

Which LLM providers does neo4j-graphrag support for entity extraction?▼

neo4j-graphrag supports OpenAI, Anthropic Claude, Vertex AI Gemini, Amazon Bedrock, Ollama for local models, and MistralAI. Each installs via an extras package such as neo4j-graphrag[openai] and shares the same LLM interface.

Why does my extraction produce duplicate entity nodes in Neo4j?▼

Duplicates appear when entity resolution is not run after ingestion. Run SinglePropertyExactMatchResolver for exact name matches or FuzzyMatchResolver for typos and alternate spellings after bulk ingest completes.

Can I import CSV or relational data with this approach?▼

No, this skill covers unstructured and semi-structured documents only. Structured CSV and relational imports should use LOAD CSV or dedicated relational import tooling instead of LLM-based extraction.

What chunk size should I use for PDF ingestion with GPT-4o?▼

Use FixedSizeSplitter with chunk_size between 256 and 512 tokens and overlap around 10 percent for GPT-4o extraction. Never exceed 2048 tokens, and stay within the embedding model limit of 8191 tokens for text-embedding-3-small.

Related Skills

Explore More Software Engineering Skills→
masermediagroup-stackmasermediagroup-stack
1

ingesting-data

Implements data ingestion pipelines from cloud storage, APIs, files, and streaming sources into databases.

Community
Advanced
eklyukineklyukin

neo4j-graphrag-skill

Build GraphRAG retrieval pipelines on Neo4j using the neo4j-graphrag Python package.

Community
Advanced
eklyukineklyukin

neo4j-document-import-skill

Ingest unstructured documents into Neo4j as a knowledge graph with LLM entity extraction.

Community
Advanced
organvm-i-theoriaorganvm-i-theoria

data-ingestion-pipeline

Build batch and streaming data ingestion pipelines with extraction, validation, deduplication, and staging.

Official
Intermediate