neo4j-graphrag-skill
installsforks
How to install neo4j-graphrag-skill?
neo4j-contrib/neo4j-skills/neo4j-graphrag-skill
$npx -y skills add neo4j-contrib/neo4j-skills --skill neo4j-graphrag-skillBuild GraphRAG retrieval pipelines on Neo4j using the neo4j-graphrag Python
SKILL.md
| 1 | # Neo4j GraphRAG Skill |
| 2 | |
| 3 | ## When to Use |
| 4 | |
| 5 | - Building GraphRAG retrieval pipelines with `neo4j-graphrag` Python package |
| 6 | - Choosing between VectorRetriever, HybridRetriever, VectorCypherRetriever, HybridCypherRetriever |
| 7 | - Writing `retrieval_query` Cypher fragments for graph-augmented context |
| 8 | - Wiring retriever + LLM into a `GraphRAG` pipeline |
| 9 | - Using LLM-routed multi-retriever with `ToolsRetriever` |
| 10 | - Debugging low retrieval quality |
| 11 | - Integrating Neo4j with LangChain, LlamaIndex, or Haystack |
| 12 | |
| 13 | ## When NOT to Use |
| 14 | |
| 15 | - **KG construction from documents** → `neo4j-document-import-skill` |
| 16 | - **Plain vector/semantic search without graph traversal** → `neo4j-vector-index-skill` |
| 17 | - **Hybrid search that combines vector with fulltext or other ranked sources** → `neo4j-vector-index-skill` |
| 18 | - **GDS algorithms (PageRank, Louvain, node embeddings)** → `neo4j-gds-skill` |
| 19 | - **Agent long-term memory** → `neo4j-agent-memory-skill` |
| 20 | - **Writing raw Cypher queries** → `neo4j-cypher-skill` |
| 21 | |
| 22 | --- |
| 23 | |
| 24 | ## Retriever Selection |
| 25 | |
| 26 | ``` |
| 27 | Has fulltext index? |
| 28 | YES → Hybrid variants (HybridRetriever / HybridCypherRetriever) |
| 29 | NO → Vector variants (VectorRetriever / VectorCypherRetriever) |
| 30 | |
| 31 | Need graph traversal after vector lookup? |
| 32 | YES → Cypher variants (VectorCypherRetriever / HybridCypherRetriever) |
| 33 | NO → plain variants |
| 34 | |
| 35 | Natural-language-to-Cypher? → Text2CypherRetriever (no embedder needed) |
| 36 | LLM should route between retrievers? → ToolsRetriever |
| 37 | Vectors stored in external DB? → WeaviateNeo4jRetriever / PineconeNeo4jRetriever / QdrantNeo4jRetriever |
| 38 | ``` |
| 39 | |
| 40 | | Retriever | Vector | Fulltext | Graph | Best For | |
| 41 | |---|:---:|:---:|:---:|---| |
| 42 | | `VectorRetriever` | ✓ | — | — | Baseline semantic search | |
| 43 | | `HybridRetriever` | ✓ | ✓ | — | Better recall, no graph expansion | |
| 44 | | `VectorCypherRetriever` | ✓ | — | ✓ | GraphRAG without fulltext | |
| 45 | | `HybridCypherRetriever` | ✓ | ✓ | ✓ | **Production GraphRAG — default** | |
| 46 | | `Text2CypherRetriever` | — | — | ✓ | NL→Cypher, no embedder | |
| 47 | | `ToolsRetriever` | varies | varies | varies | LLM-routed multi-retriever | |
| 48 | | `WeaviateNeo4jRetriever` | ✓ | — | ✓ | Vectors in Weaviate | |
| 49 | | `PineconeNeo4jRetriever` | ✓ | — | ✓ | Vectors in Pinecone | |
| 50 | | `QdrantNeo4jRetriever` | ✓ | — | ✓ | Vectors in Qdrant | |
| 51 | |
| 52 | --- |
| 53 | |
| 54 | ## Install |
| 55 | |
| 56 | ```bash |
| 57 | pip install neo4j-graphrag[openai] # OpenAI LLM + embeddings |
| 58 | pip install neo4j-graphrag[anthropic] # Anthropic Claude |
| 59 | pip install neo4j-graphrag[google] # Vertex AI / Gemini |
| 60 | pip install neo4j-graphrag[bedrock] # Amazon Bedrock (boto3) |
| 61 | pip install neo4j-graphrag[cohere] # Cohere |
| 62 | pip install neo4j-graphrag[mistralai] # MistralAI |
| 63 | pip install neo4j-graphrag[ollama] # Ollama (local) |
| 64 | pip install neo4j-graphrag[weaviate] # Weaviate external retriever |
| 65 | pip install neo4j-graphrag[pinecone] # Pinecone external retriever |
| 66 | pip install neo4j-graphrag[qdrant] # Qdrant external retriever |
| 67 | ``` |
| 68 | |
| 69 | Requires: Python >= 3.10, `neo4j >= 5.17.0` (driver 6.x supported). |
| 70 | |
| 71 | --- |
| 72 | |
| 73 | ## Step 2 — Choose Retriever |
| 74 | |
| 75 | ``` |
| 76 | Has fulltext index? YES → Hybrid variants (better recall) |
| 77 | NO → Vector variants (baseline) |
| 78 | |
| 79 | Needs graph context after vector lookup? YES → Cypher variants |
| 80 | NO → plain variants |
| 81 | |
| 82 | For natural-language-to-Cypher? → Text2CypherRetriever (no embedder needed) |
| 83 | For multi-tool LLM routing? → ToolsRetriever |
| 84 | Using external vector DB? → WeaviateNeo4jRetriever / PineconeNeo4jRetriever / QdrantNeo4jRetriever |
| 85 | ``` |
| 86 | |
| 87 | | Retriever | Vector | Fulltext | Graph | When to use | |
| 88 | |---|:---:|:---:|:---:|---| |
| 89 | | `VectorRetriever` | ✓ | — | — | Baseline; quick start | |
| 90 | | `HybridRetriever` | ✓ | ✓ | — | Better recall; no graph context | |
| 91 | | `VectorCypherRetriever` | ✓ | — | ✓ | GraphRAG without fulltext | |
| 92 | | `HybridCypherRetriever` | ✓ | ✓ | ✓ | **Production GraphRAG — default choice** | |
| 93 | | `Text2CypherRetriever` | — | — | ✓ | LLM generates Cypher; no embedder | |
| 94 | | `ToolsRetriever` | varies | varies | varies | Multi-retriever LLM routing | |
| 95 | |
| 96 | For custom Cypher hybrid |
People who liked this also viewed
6 picksType
- azure-storageAzure Storage Services including Blob Storage, File Shares, Queue Storage, Table Storage, and Data Lake. Answers questions about storage access tiers (hot, cool, cold, archive), when to use each tier,SkillsJul 2026
- supabase-postgres-best-practicesPostgres performance optimization and best practices from Supabase. Use this skill when writing, reviewing, or optimizing Postgres queries, schema designs, or database configurations.SkillsJul 2026
- supabaseUse when doing ANY task involving Supabase. Triggers: Supabase products (Database, Auth, Edge Functions, Realtime, Storage, Vectors, Cron, Queues); client libraries and SSR integrations (supabase-js, SkillsJul 2026
- convex-quickstartCreates or adds Convex to an app. Use for new Convex projects, npm create convex@latest, frontend setup, env vars, or the first npx convex dev run.SkillsJul 2026
- convex-performance-auditAudits Convex performance for reads, subscriptions, write contention, and function limits. Use for slow features, insights findings, OCC conflicts, or read amplification.SkillsJul 2026
- convex-setup-authSets up Convex auth, identity mapping, and access control. Use for login, auth providers, users tables, protected functions, or roles in a Convex app.SkillsJul 2026
