PlugWiki
amazon-neptune-mcp-server logo

Amazon Neptune MCP Server: Setup for Claude, Cursor and Other Apps

Official · by Amazon Web ServicesLocal
GetPyPIawslabs.amazon-neptune-mcp-serverawslabs/mcp/tree/HEAD/src/amazon-neptune-mcp-server

Lets your AI assistant look at an Amazon Neptune graph and run openCypher or Gremlin queries on it.

This is an AWS Labs server that runs on your computer and connects to one Neptune graph that you name in the setup. The assistant can check the graph status, read its schema and run openCypher or Gremlin queries.

The server runs any query it is given, including ones that change or delete data. Use an AWS role that allows only the read actions you want. Your computer must also be able to reach the graph, which for Neptune Database sits in a private network.

Good for

  • Developers with data in Amazon Neptune who want to explore it by asking questions in plain words
  • Analysts who want the assistant to write and run graph queries for them
  • Teams on Neptune Analytics or Neptune Database that can reach the graph from their own machine

Works in

12 should work

Tap an app to see how to set it up and any known problems.

Install amazon-neptune-mcp-server

Claude Code

  1. 1

    Copy this prompt.

    Add the amazon-neptune-mcp-server server to Claude Code. Run this command:
    
    claude mcp add amazon-neptune -e NEPTUNE_ENDPOINT=<NEPTUNE_ENDPOINT> -- uvx awslabs.amazon-neptune-mcp-server
    
    Ask me for NEPTUNE_ENDPOINT and only put the values in that config, nowhere else.
    
    Then check that it loaded and tell me if I need to sign in or restart.
  2. 2

    Paste it into Claude Code. It shows the change and asks before running it.

Requires uv (Python).

Built from the package data we hold. Not yet tested by us.

Tools

  • Runs an openCypher query on your graph

    Runs the openCypher query you give it against the graph set in your setup and returns the result. It works on both Neptune Database and Neptune Analytics, and it can change data if the query does.

    Try: “Using the Neptune graph, find the ten people with the most connections.”

    Delete · run_opencypher_query

  • Runs a Gremlin query on your graph

    Runs the Gremlin query you give it against the graph. The README lists Gremlin for Neptune Database only, and the query can change data.

    Try: “Run a Gremlin query that counts vertices by label.”

    Delete · run_gremlin_query

  • Shows the labels and links in your graph

    Returns the graph schema: the vertex and edge labels and which vertices connect through which edges. The assistant uses it to write correct queries.

    Try: “What does my Neptune graph contain? Show me the schema.”

    Network · get_graph_schema

  • Checks whether your graph is available

    Reports whether the configured graph is Available or Unavailable to the server. Use it to confirm the connection works.

    Try: “Is my Neptune graph reachable right now?”

    Network · get_graph_status

Read from the source code

Permissions

high risk · 3 things to watch

High risk

The assistant can run any query on your graph, including ones that change or delete data, with the permissions of your AWS role. Use a role that allows only read queries.

What it can do

  • Read your data
  • Create and change things
  • Delete things
  • Read-only mode

Watch out for

  • It runs any query, including ones that delete data

    The README warns that the server runs whatever query it receives. Use an AWS role that allows only read queries unless you want changes. github.com

  • Query results go to the assistant

    Whatever the graph returns is sent to the assistant and its provider. Keep sensitive graphs out of reach of the role you give it. github.com

  • Your computer needs network access to the graph

    Neptune Database sits in a private network, so you need a route into it. Neptune Analytics can be reached through a public endpoint if you set one up. github.com

Security advisories

No published security advisories as of Nov 1, 2026.

Technical details
Turning on read-onlyThe server has no read-only switch in the README. It says to limit the AWS role to the data actions you allow, as described in the Neptune Database and Neptune Analytics security guides.
Talks toYour Neptune graph endpoint on port 8182
Sign-in and keysUses your AWS credentials from AWS_PROFILE or environment variables. The server keeps no separate account.

Questions

Can it change or delete data in my graph?

Yes. It runs whatever openCypher or Gremlin query it is given, so a query can change or delete data if your AWS role allows it. Use a read-only role to prevent that.

What format does the endpoint take?

For Neptune Database use neptune-db://<cluster endpoint>. For Neptune Analytics use neptune-graph://<graph identifier>. Set it in the NEPTUNE_ENDPOINT setting in the install box.

Which query languages does it support?

openCypher and Gremlin on Neptune Database, and openCypher on Neptune Analytics.

Why can't it reach my Neptune database?

Neptune Database lives in a private network, so the computer running the server needs a route into it, such as a VPN or a machine inside the network. Neptune Analytics can also be reached through a public endpoint if you set one up.

Can I run it with Docker?

Only after you build the image yourself, as the README shows. The command in the install box is simpler.

Can it run read-only?

Not through a server setting. The README says to restrict the AWS role to the actions you want to allow.

Information

updated Nov 2, 2026

Publisher
Amazon Web Services (official)
Package
awslabs.amazon-neptune-mcp-server
Registry
PyPI
Transport
stdio
Auth
token in env
License
Apache-2.0
Last commit
Oct 30, 2026
Open issues
143
Listed in
Docker MCP Catalog

GitHubDockerDocker MCP Catalog

Other AWS MCP servers

Sources

Docs, issue trackers and discussions this page is built from. Issues, quotes and findings above link to their own source.

Show all 2
  1. Amazon Neptune MCP Server README · github.com
  2. Neptune server source · github.com

Facts from the official MCP registry, GitHub, npm, PyPI and the Docker catalog. Updated Nov 2, 2026.