Official Neo4j MCP server for giving Claude structured access to Neo4j graph schema introspection, read-only Cypher, optional write Cypher, and Graph Data Science procedure discovery.
Neo4j MCP can expose graph schema, labels, relationship types, property keys, nodes, relationships, paths, and procedure availability., The read-cypher tool is documented as read-only and uses EXPLAIN plus Neo4j query-type classification, but custom procedures or functions incorrectly classified as read-only may bypass that check., The write-cypher tool can execute database mutations and is disabled when NEO4J_READ_ONLY is true., GDS procedure discovery can reveal installed analytics capabilities and graph-processing surface area., Connecting to production graph databases should use least-privilege credentials, read-only mode by default, and human approval for any write query., Telemetry is configurable with `NEO4J_TELEMETRY`; teams should decide whether to disable it before sensitive use.
Privacy notes
Graph schema, sampled properties, returned nodes, relationships, paths, query text, and GDS procedure lists can expose customer data, internal topology, access patterns, business logic, or sensitive relationships., NEO4J_URI, NEO4J_USERNAME, NEO4J_PASSWORD, database names, Aura details, and query logs should stay out of prompts, issues, logs, screenshots, and committed files., Claude outputs based on graph data may re-expose retrieved private data in transcripts or downstream tickets., Log format, log level, schema sample size, and telemetry settings should be reviewed before connecting regulated or production datasets.
Author
Neo4j
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-06
Decision playbook
Review trust signals before you adopt
Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.
Compare context
Selected
0
Current score
63
Baseline
—
Delta
No baseline selected
No major trust-signal divergence detected in the current selection.
Source and provenance checks
Needs review
Confirm ownership and provenance before trusting install instructions.
Source link availableRequired
Open the canonical repository and verify ownership.
Done
Source provenance statusRequired
Marked as source-backed.
Done
Metadata reviewed
No reviewed flag detected in metadata.
Pending
Safety and privacy checks
Complete
Validate risk disclosures before installation or API wiring.
Safety notes presentRequired
Review the listed safety guidance before running commands.
Done
Privacy notes presentRequired
Review data handling notes before connecting accounts or secrets.
Done
Trust level risk gateRequired
Trust level does not block evaluation.
Done
Package and install checks
Needs review
Check package metadata and artifact integrity signals.
Install payload available
Install or copy payload is available for review.
Done
Package verification flag
No package verification flag provided.
Pending
Checksum metadata
No checksum provided for downloaded artifact.
Pending
Compare-driven decision checks
Needs review
Use compare context to validate trade-offs before adoption.
Compare tray has multiple entries
Add at least one more entry to compare trust differences.
6 safety and 4 privacy notes across 6 risk areas. Review closely: credentials & tokens.
6 areas
SafetyLocal filesNeo4j MCP can expose graph schema, labels, relationship types, property keys, nodes, relationships, paths, and procedure availability.
SafetyGeneralThe read-cypher tool is documented as read-only and uses EXPLAIN plus Neo4j query-type classification, but custom procedures or functions incorrectly classified as read-only may bypass that check.
SafetyExecution & processesThe write-cypher tool can execute database mutations and is disabled when NEO4J_READ_ONLY is true.
SafetyExecution & processesGDS procedure discovery can reveal installed analytics capabilities and graph-processing surface area.
SafetyCredentials & tokensConnecting to production graph databases should use least-privilege credentials, read-only mode by default, and human approval for any write query.
SafetyTelemetryTelemetry is configurable with `NEO4J_TELEMETRY`; teams should decide whether to disable it before sensitive use.
PrivacyLocal filesGraph schema, sampled properties, returned nodes, relationships, paths, query text, and GDS procedure lists can expose customer data, internal topology, access patterns, business logic, or sensitive relationships.
PrivacyCredentials & tokensNEO4J_URI, NEO4J_USERNAME, NEO4J_PASSWORD, database names, Aura details, and query logs should stay out of prompts, issues, logs, screenshots, and committed files.
PrivacyExecution & processesClaude outputs based on graph data may re-expose retrieved private data in transcripts or downstream tickets.
PrivacyData retentionLog format, log level, schema sample size, and telemetry settings should be reviewed before connecting regulated or production datasets.
Safety notes
Neo4j MCP can expose graph schema, labels, relationship types, property keys, nodes, relationships, paths, and procedure availability.
The read-cypher tool is documented as read-only and uses EXPLAIN plus Neo4j query-type classification, but custom procedures or functions incorrectly classified as read-only may bypass that check.
The write-cypher tool can execute database mutations and is disabled when NEO4J_READ_ONLY is true.
GDS procedure discovery can reveal installed analytics capabilities and graph-processing surface area.
Connecting to production graph databases should use least-privilege credentials, read-only mode by default, and human approval for any write query.
Telemetry is configurable with `NEO4J_TELEMETRY`; teams should decide whether to disable it before sensitive use.
Privacy notes
Graph schema, sampled properties, returned nodes, relationships, paths, query text, and GDS procedure lists can expose customer data, internal topology, access patterns, business logic, or sensitive relationships.
NEO4J_URI, NEO4J_USERNAME, NEO4J_PASSWORD, database names, Aura details, and query logs should stay out of prompts, issues, logs, screenshots, and committed files.
Claude outputs based on graph data may re-expose retrieved private data in transcripts or downstream tickets.
Log format, log level, schema sample size, and telemetry settings should be reviewed before connecting regulated or production datasets.
Prerequisites
Python runtime and neo4j-mcp-server package installed for stdio usage.
Neo4j database URI, username, password, and database name.
Read-only mode enabled unless the workflow explicitly needs write Cypher.
Least-privilege Neo4j user or role scoped to the database and operations Claude should perform.
Review of telemetry, log level, schema sample size, and graph data sensitivity before connecting production databases.
Neo4j MCP Server is the official Model Context Protocol server for Neo4j. It
lets Claude and other MCP clients inspect graph schema, execute read-only
Cypher, optionally execute write Cypher, and list available Graph Data Science
procedures against a configured Neo4j database.
The official server lives in neo4j/mcp and is separate from the Neo4j Labs MCP
server bundle in neo4j-contrib/mcp-neo4j. The Labs bundle contains
experimental servers, while this entry is scoped to Neo4j's official product
MCP server.
These sources were reviewed on 2026-06-06. Prefer the official Neo4j MCP
docs, repository README, package metadata, manifest, license, and notice file
for current install commands, supported tools, configuration, telemetry, and
licensing.
Features
Official Neo4j MCP server from neo4j/mcp.
PyPI package neo4j-mcp-server.
Stdio configuration for Python module launch.
get-schema tool for labels, relationship types, and property keys.
read-cypher tool for read-only Cypher queries.
write-cypher tool for write queries when read-only mode is not enabled.
list-gds-procedures tool for Graph Data Science procedure discovery.
Environment configuration for URI, username, password, database, read-only mode, telemetry, logs, and schema sampling.
Manifest metadata for packaged MCPB-style installs.
Documentation for installation and configuration in Neo4j's official docs.
Installation
Install the package:
pip install neo4j-mcp-server
Then configure an MCP client with Neo4j credentials. For safer browsing, start
with read-only mode enabled:
Ask Claude to inspect graph schema before writing Cypher.
Run read-only Cypher to answer graph questions from nodes, relationships, and paths.
Generate graph summaries from a bounded query result.
Explore which Graph Data Science procedures are available in the instance.
Compare schema changes before and after a migration.
Keep write mode disabled for analyst workflows and enable it only for reviewed maintenance tasks.
Use least-privilege Neo4j credentials for a single approved database.
Safety and Privacy
Neo4j MCP connects Claude to a graph database, so schema inspection alone can
reveal sensitive business structure. Query results can expose private
relationships, customer records, identity data, infrastructure topology,
fraud-investigation graphs, or other high-context data.
Use read-only mode by default, scope credentials tightly, and require human
approval for any write query. Review generated Cypher before execution, and be
careful with custom procedures or functions because the README documents that
misclassified read-only procedures can bypass the read-only check.
Disclosure
Neo4j offers open-source and commercial graph database products and Aura cloud
services. This listing is not sponsored, paid, or affiliate-driven, and it is
scoped to the official source-backed MCP server.
Duplicate Check
No Neo4j MCP Server, neo4j/mcp, neo4j-mcp-server, or official Neo4j MCP
entry was found in content/mcp, content/tools, content/guides,
content/agents, or content/skills. Existing mentions of Neo4j are
incidental examples in other database and graph-related entries.
Show that Neo4j MCP Server is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.
[](https://heyclau.de/entry/mcp/neo4j-mcp-server)
How it compares
Neo4j MCP Server side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
3 trust signals differ across this comparison (Package trust, Source provenance, Submitter).
Official Neo4j MCP server for giving Claude structured access to Neo4j graph schema introspection, read-only Cypher, optional write Cypher, and Graph Data Science procedure discovery.
Official reference MCP server that gives Claude persistent memory through a local knowledge graph of entities, relations, and observations, stored in a file on your own machine.
✓Neo4j MCP can expose graph schema, labels, relationship types, property keys, nodes, relationships, paths, and procedure availability.
The read-cypher tool is documented as read-only and uses EXPLAIN plus Neo4j query-type classification, but custom procedures or functions incorrectly classified as read-only may bypass that check.
The write-cypher tool can execute database mutations and is disabled when NEO4J_READ_ONLY is true.
GDS procedure discovery can reveal installed analytics capabilities and graph-processing surface area.
Connecting to production graph databases should use least-privilege credentials, read-only mode by default, and human approval for any write query.
Telemetry is configurable with `NEO4J_TELEMETRY`; teams should decide whether to disable it before sensitive use.
✓MariaDB MCP connects Claude to a live database and exposes schema inspection plus SQL execution tools.
The default read-only mode allows SELECT, SHOW, DESCRIBE, DESC, and USE-style queries, but the README warns that database privileges are the only reliable way to guarantee read-only access.
The server includes a `create_database` tool and optional vector-store tools that can create, insert into, search, and delete vector-store tables when enabled.
If `MCP_READ_ONLY=false` or credentials have broad privileges, model-generated SQL can create, modify, delete, or expose database state.
The server checks for risky FILE privilege behavior, but teams should revoke FILE and other unnecessary global privileges from the connected MariaDB user.
HTTP and SSE transports require explicit authentication and restricted host/origin settings before any non-local use.
Embedding providers can receive document text or derived content when vector-store tooling is enabled.
✓Persists remembered information to a local file on disk (its location can be set with MEMORY_FILE_PATH); manage and review that file as it accumulates data.
Runs locally as an stdio process and makes no external network calls; the knowledge graph stays on your machine.
✓Use a read-only database user with restricted schemas unless write access is explicitly needed and reviewed.
Privacy notes
✓Graph schema, sampled properties, returned nodes, relationships, paths, query text, and GDS procedure lists can expose customer data, internal topology, access patterns, business logic, or sensitive relationships.
NEO4J_URI, NEO4J_USERNAME, NEO4J_PASSWORD, database names, Aura details, and query logs should stay out of prompts, issues, logs, screenshots, and committed files.
Claude outputs based on graph data may re-expose retrieved private data in transcripts or downstream tickets.
Log format, log level, schema sample size, and telemetry settings should be reviewed before connecting regulated or production datasets.
✓Database credentials, hostnames, database names, schemas, table names, column names, SQL text, query results, errors, and log files may be visible to the MCP client and model provider.
MariaDB data can include customer records, credentials, audit logs, business metrics, payment data, healthcare data, or other regulated information.
Vector-store tables can persist source documents, embeddings, metadata, and semantic-search results inside MariaDB.
OPENAI_API_KEY, GEMINI_API_KEY, HF_MODEL settings, SSL certificate paths, dotenv files, and database passwords should stay out of prompts, issues, logs, screenshots, and committed files.
Review log retention because the server writes logs to `logs/mcp_server.log` by default.
✓Details Claude chooses to remember are written to and read from the local memory file, so it can retain personal or project information across sessions.
Review or clear the memory file if it should not retain sensitive data, and use MEMORY_FILE_PATH to control where it is stored.
✓Queried schemas, table names, rows, connection details, and application data may be exposed through tool calls.
Prerequisites
Python runtime and neo4j-mcp-server package installed for stdio usage.
Neo4j database URI, username, password, and database name.
Read-only mode enabled unless the workflow explicitly needs write Cypher.
Least-privilege Neo4j user or role scoped to the database and operations Claude should perform.
Python 3.11 and uv available to the MCP client runtime.
Reviewed checkout of the MariaDB MCP repository with dependencies installed.
MariaDB database host, port, database name, username, and password.
Least-privilege MariaDB user scoped to only the databases and operations Claude should access.
Node.js 18+ and npx available (verify with: npx --version)
Claude Code or Claude Desktop with MCP support
A writable location for the memory file, optionally set with the MEMORY_FILE_PATH environment variable
PostgreSQL database (local or remote PostgreSQL server)
PostgreSQL connection string (PostgreSQL connection URI with user, password, host, port, and database format)
Node.js and npx (for running @modelcontextprotocol/server-postgres package)
Network access (if connecting to remote PostgreSQL server)
Install
pip install neo4j-mcp-server
uv run server.py
claude mcp add memory -- npx -y @modelcontextprotocol/server-memory